323 1 DIVISION OF ADMINISTRATIVE HEARINGS DEPARTMENT OF ADMINISTRATION, STATE OF FLORIDA 2 SUGAR CANE GROWERS COOPERATIVE ) 3 OF FLORIDA; ROTH FARMS, INC.; and ) WEDGWORTH FARMS, INC., ) 4 Petitioners, ) vs. )DOAH Case No. 92-3038 5 SOUTH FLORIDA WATER MANAGEMENT ) 92-3039 DISTRICT, an agency of the State ) 92-3040 6 of Florida; et al., ) 92-6796 Respondents. ) 92-6797 7 - - - - - - - - - - - - - - - - - x 92-6799 FLORIDA SUGAR CANE LEAGUE, INC., ) 92-6800 8 UNITED STATES SUGAR CORPORATION; ) and NEW HOPE SOUTH, INC., ) 9 Petitioners, ) vs. ) 10 SOUTH FLORIDA WATER MANAGEMENT ) DISTRICT, an agency of the State ) 11 of Florida; et al., ) Respondents. ) 12 - - - - - - - - - - - - - - - - - x FLORIDA FRUIT AND VEGETABLE ) 13 ASSOCIATION; LEWIS POPE FARMS; ) W.E. SCHLECHTER & SONS, INC., ) 14 and HUNDLEY FARMS, INC., ) Petitioners, ) 15 vs. ) SOUTH FLORIDA WATER MANAGEMENT ) 16 DISTRICT, an agency of the State ) of Florida; et al., ) 17 Respondents. ) - - - - - - - - - - - - - - - - - x 18 100 Southeast 2nd Street Miami, Florida 19 March 2nd, 1994 1:10 p.m.- 5:35 p.m. 20 DEPOSITION OF DENNIS P. LETTENMAIER 21 VOLUME IV 22 Taken before BARNET I ABRAMOWITZ, court 23 reporter and Notary Public in and for the State of 24 Florida at Large, pursuant to Notice of Taking 25 Deposition filed in the above cause. 324 1 AFTERNOON SESSION 2 1:10 p.m. 3 BY MR. LOREDO: 4 Q. Doctor, before lunch, you were talking 5 about -- 6 A. We are presumably on the record and all 7 that good stuff? 8 Q. On the record and you are still under 9 oath. 10 A. All right. 11 Q. Prior to lunch, we were talking about the 12 report which is dated January 5th, 1993, the Review 13 of Kriging Methods Applied to Estimation of Soil 14 Phosphorus in WCA-2A. 15 This was your initial report as to your 16 analysis as to the soil phosphorus in WCA-2A? 17 A. Yes. 18 Q. I think it might be better to work 19 backwards, and this way we can figure out what values 20 your earlier reports have now and work now with your 21 most recent reports. 22 In doing so, I turn your attention then to 23 the report dated October 15th, 1993 titled: 24 Estimation of WCA-2A Sediment Phosphorus Deposition 25 Using Reddy and Richardson Dated Core Data. 325 1 If you could take me, like you did in the 2 original report, and give me a basic overview of it, 3 what your conclusions were and the basis for 4 conclusions, and what your reliance on this report is 5 for the hearing, we can proceed that way. 6 A. Well, this report just does a spacial 7 analysis of the dated core data based on sources that 8 are documented in here which are Richardson and 9 Reddy, and the data again is theirs and is described 10 in reports by them, so I don't particularly want to 11 go into that. 12 But the idea is that they are able to 13 identify, based on cesium -- C E S I U M -- dating, 14 they can identify 1964 which essentially was the date 15 of the Test Ban Treaty. 16 And from that, one can then go estimate a 17 deposition rate or how much phosphorus has been 18 deposited in the WCA-2A to which this report applies 19 in that period of 20-some -- now almost 30 years. 20 So we simply have taken their data and 21 applied the same kriging analysis to that, that we 22 did to the other fixed depth data, which is what was 23 in the earlier reports we discussed. 24 Q. But it wasn't the same, maybe the same 25 kriging analysis, but a different program now? 326 1 A. Oh, the actual program that was used? 2 Q. Yes. 3 A. Yes, I believe it is GEOEAS in that case. 4 Q. Now, cesium data, it is called? Can you 5 pronounce that again for me? 6 A. Cesium. 7 Q. What is that? 8 A. I'm not a chemist, you really need to go 9 read their reports which describe that method. We 10 are simply using their numbers. 11 I can give you a cursory understanding? 12 Q. If you would. 13 A. Atmospheric fallout -- cesium 137, the 14 primary source, as I understand it, was weapons 15 testing, atmospheric weapons testing. So they can go 16 back through their core and look for a signal at each 17 depth of how much cesium 137 is there, and since that 18 stopped in 1964, they know by looking at a depth 19 profile in the core where 1964 was. 20 So presumably all the phosphorus in the 21 core from that point on up is what has been deposited 22 since 1964 so they can estimate a rate. 23 Q. What did you conclude from your analysis? 24 A. I think conclusion isn't maybe the proper 25 word. We did an analysis, the results of which are 327 1 maps of the mass of phosphorus that had been 2 deposited. 3 Q. Since 1964? 4 A. Yes. Well, if you divide by the years, 5 you get a rate, but -- 6 Q. Is that assuming a constant rate then per 7 year? 8 A. Well, if you divide out by the number of 9 years, yes, but again that's all somebody else's 10 analysis and you can go quibble about -- you can 11 quibble about that from the reports of either 12 Richardson or Reddy. Our analysis was simply to map 13 that out. 14 Q. Did you map out something different than 15 -- well, did Reddy and Richardson map anything out? 16 A. No, to my knowledge, no, the spacial 17 analysis had not been done because neither of them 18 had sufficient samples in and of themselves to do 19 that. 20 Q. Was your analysis consistent with theirs, 21 Richardson and Reddy? Maybe that's not a proper 22 question? 23 A. That's not quite a relevant question. 24 Q. You are saying the results of your work 25 are the maps set forth in figures 1 through 5 of this 328 1 report? 2 A. That's correct. 3 Q. And that's a result of using the kriging 4 with the GEOEAS program? 5 A. Right, applied to their data. 6 Q. What is the significance of the results 7 here? 8 A. Well, it simply allows you to go, to have 9 a spacial estimate as opposed to numbers at 10 particular locations. 11 Q. Spacial estimates? 12 A. Yes, that's what kriging does, it gives 13 you a map. 14 Q. Of the observed points? 15 A. Yes, whereas their report dealt with the 16 values at the particular points where they extracted 17 their cores. 18 Q. Are we talking about phosphorus levels at 19 different points in a depth or are we talking about 20 over time? 21 A. No, there are cores that are taken at one 22 particular time. 23 Q. Did you test your analysis and whether the 24 observed actual points came up using GEOEAS or 25 GEOPACK as you did using the SURFER program? 329 1 A. In this analysis? GEOEAS was used here, 2 not GEOPACK. 3 Q. I'm sorry, GEOEAS. 4 What I was trying to get is to when we 5 started talking about the first report, the SURFER, 6 you said in your analysis, it turned out, you 7 believe, there was a smoothing process, and you did 8 that by checking, and I guess maybe redoing the work 9 or reinserting the data and asking the program to 10 give you the points or give you the observed point, 11 and you said it did not give you the observed point 12 using the SURFER program. 13 Therefore, you assumed or you stated that 14 most likely it smoothed out, there was smoothing out 15 in the program. 16 Do you understand what I'm talking about? 17 A. Yes. 18 Q. I am asking if you did that same thing 19 with GEOEAS? 20 A. We verified that GEOEAS does reproduce. 21 Q. Did you verify GEOEAS? 22 A. Yes. 23 Q. What were the results of that? 24 A. If you are asking me to recall what 25 particular data set we used and so on, I don't recall 330 1 that, but we did do a test to make sure that that was 2 in fact happening. 3 Q. And it passed your test? 4 A. Yes. 5 Q. Meaning it gave you the observed points? 6 A. Well, mathematically it has to. The only 7 question is if something else is being done to 8 post-process the data, and it was our conclusion -- 9 and one would not expect the data to be 10 post-processed unless you asked it to, and it was 11 not. 12 Q. So the program itself did not ask to do 13 it. So there was no way that it should not have 14 passed the test or the data verified in the GEOEAS 15 program? 16 A. GEOEAS does do kriging and kriging alone, 17 and kriging has properties that at observed points 18 are reproduced, and our tests showed it was working 19 essentially as advertised. 20 Q. Did you test it only once? 21 A. I don't understand the relevance of that 22 question. 23 Q. Well, when you test something, you can 24 either do it once or 10 times. 25 A. If you multiply two times 2 in the 331 1 computer 10 times, it will always come out 4. At 2 some point, one does a verification of something and 3 it either works or it doesn't. This is not like a 4 laboratory analysis where one is testing for it to be 5 in control or out of control. 6 Q. Now, you state in figure 1 of the report, 7 you state fictitious sampling locations. 8 Were those points that you are asking 9 kriging analysis to pick out and give you what data 10 should be there? 11 A. No. 12 Q. What are the fictitious points for? 13 A. One of the properties of kriging is that 14 it gives you -- in addition to the best estimate, it 15 gives you the variance or uncertainty of that 16 estimate. 17 The property of kriging is that that 18 variance does not depend on the value of the 19 observation, it depends only on the variogram and on 20 the locations at which the samples are collected. 21 So commonly, common use for sampling 22 design of kriging is to ask questions: if you were to 23 take additional samples, where would you take them, 24 where would they have the most value? 25 So essentially one can trick the software 332 1 into giving you that answer by inserting fictitious 2 sampling locations and see how they reduce the 3 variance, and one can put any numerical value because 4 it will not affect the variance. 5 The value one puts in at any fictitious 6 sampling location will not affect the variance, only 7 the location, so one can test at different locations 8 how much influence they might have on the variance. 9 Q. Do you control the variogram in that type 10 of analysis? Are you putting inputs or making 11 assumptions in the relationships or in the variance 12 in this type of program? 13 A. Well, the variogram can be either 14 estimated internally in the program or specified. 15 Q. And you did? 16 A. In that particular report, we specified 17 it. 18 Q. What was the other way? 19 A. Well, it can be estimated directly from 20 the data. 21 Q. What would be the intended use of this 22 report for the Park? 23 A. The estimated deposition of phosphorus in 24 WCA-2A. 25 Q. Is there another way to do it? 333 1 A. To do what? 2 Q. To estimate the phosphorus in the WCA-2A? 3 A. At any arbitrary point? 4 Q. Yes. 5 A. I'm sure one can dream up an infinity of 6 ways of doing it. 7 Q. Do you know of a better or worse way, a 8 better way? 9 A. I can give you a worse way. 10 Q. What's the worst way? 11 A. Simply go average all the values. 12 Q. What's the best way? 13 A. This way is certainly as defensible, 14 certainly as defensible as any. It makes in my 15 judgment the best use of the available data. 16 Q. Have you seen any other similar type of 17 analysis anywhere else other than the Everglades? 18 A. Use of kriging? 19 Q. Not kriging, but in a core analysis? 20 A. Core analysis? 21 Q. Yes. 22 A. I'm not particularly familiar with core 23 analysis, no. 24 Q. That's not the question I wanted to ask. 25 What we are trying to do here is develop 334 1 or understand what phosphorus is in areas where there 2 wasn't data collected. Is that what you have done? 3 A. I don't understand that question. 4 Q. Kriging, I thought the whole thing about 5 kriging is to estimate what the results would be in 6 an unobserved area? 7 A. Yes, it is spacial interpolator. I'm not 8 quite sure what you mean by "unobserved area." There 9 are observations in the area. There are not 10 observations at all points. 11 Q. That's what I have. To be more technical, 12 I am talking about points then. It is to give you an 13 estimation of what the calculation of the phosphorus 14 would be at a particular point that had not been, I 15 guess, observed at that point, no data collected at 16 that point, and kriging develops through its analysis 17 what that particular point would be in terms of 18 phosphorus? 19 A. Yes. 20 Q. That's what you have done, you have taken 21 data that you used kriging and then created your 22 contour maps? 23 A. Yes. 24 Q. Tell me in layman's terms, the difference 25 in figure 4 and 5. 335 1 A. Oh, well, one of these is the kriging map. 2 Q. Which one? 3 A. Figure 4, and figure 5 is the standard, 4 kriging standard deviation map. 5 Q. What's the difference? 6 A. I think I indicated to you before that 7 kriging is a spacial interpolator. 8 Q. That I understand. 9 A. So one of those is the spacial 10 interpolation for the 64 to 90 values which includes 11 some specification of the fictitious points which I 12 think were indicated in another figure, and the other 13 is the variance map or standard deviation, actually 14 kriging standard deviation for that. 15 Q. They are both krigings, right? 16 A. No, one is a standard deviation which 17 essentially is the uncertainty, the other is the 18 actual estimated values. You get contours of both 19 the values and their standard deviations. 20 Q. Okay. Now I follow you. 21 Why don't we mark this as an exhibit so we 22 know what we are talking about. 23 (Lettenmaier Exhibit 6 was marked for 24 identification) 25 MR. LOREDO: If you could mark as Exhibit 336 1 7, the report we already talked about, report 1, the 2 Review of Kriging Methods as Applied to Estimation of 3 Soil Phosphorus in WCA-2A. 4 Make that Exhibit 7. 5 You already talked about this one and I 6 would like to mark this for the record. 7 (Lettenmaier's Exhibit 7 was marked for 8 identification) 9 BY MR. LOREDO: 10 Q. If I recall standard deviation, that's 11 deviation from? 12 A. Square root of the variance. 13 Q. What is it supposed to show? What is it 14 meant to reflect, the standard deviation in layman's 15 terms? 16 A. You mean a definition of the variance? 17 Mathematical definition of the variance is the 18 expectation of value minus its means squared. 19 Q. And then the square root of that? 20 A. The square root of that is the standard 21 deviation. The practical implication is that it 22 gives you -- it is related to the uncertainty of the 23 estimate. 24 Q. In percentages or? 25 A. No -- well, as plotted there, those are 337 1 actual numbers in the same units as the data, i.e., 2 if the data in the previous figure is the 3 concentration, which I think it might be that the 4 standard deviation is a concentration as well. 5 Q. Let me see if you can help me. 6 On figure 4, which is the kriging -- the 7 spacial interpolation, if I look in figure 4, I see 8 that the concentration, to use north at the top -- 9 A. Which I think it is. 10 Q. Okay, looking at maybe the northeast 11 corner, this part I see a contour of 1250 and another 12 of 1150. That's in parts per million, phosphorus 13 concentration? 14 A. I need to look at the caption there. 15 (Pause) 16 A. In parts per million, yes. 17 Q. Now, if you go to figure 5 and look in the 18 same area, the lines aren't drawn the same, but it 19 seems more running north-south, but I say around a 20 300 there. The 300, is that telling me that -- if I 21 go back to figure 4 and 1250, it is plus or minus 300 22 parts per million? 23 A. Its standard deviation is 300, plus or 24 minus is your interpretation. 25 Q. Well, is that an incorrect interpretation? 338 1 A. Well, if the probability distribution were 2 normal, one standard deviation includes something 3 around, between plus one standard deviation and minus 4 one standard deviation, is a confidence value that 5 includes probability, something like 68 percent. 6 Q. So when I say plus or minus 300, you are 7 saying that's a 68 percent confidence level? 8 A. No, I didn't say that. I said if it were 9 normally distributed. 10 Q. If it were normally distributed, my 11 statement of plus or minus 300 would be 68 percent 12 confidence level? 13 A. That's correct. 14 Q. But we can't assume a normal distribution 15 level here? 16 A. That is a correct statement. 17 Q. We can't do that because there's too many 18 variables involved? 19 A. Well, for one reason, normal distribution 20 is a symmetric distribution, but clearly a variable 21 that's bounded below by zero cannot strictly be 22 symetric. You can't have negative concentrations. 23 Q. If we go on the higher standard deviation 24 instead of 1, 2, where does the percentage then go, 25 up or down? I said plus or minus 2 standard 339 1 deviation? 2 A. Between minus 2 standard deviations and 3 plus 2 standard deviations for normal distributions, 4 contains around 95 percent probability. 5 Q. Again, you caveated that it be a normal 6 distribution? 7 A. Yes. 8 Q. I feel like I am going through statistics 9 again. 10 I think I know what you did in this 11 report, Exhibit 6. 12 Has that changed your preliminary analysis 13 of any other reports? For example, I guess I am 14 making reference now to what we have marked as 15 Exhibit 7, when you reviewed the kriging method? 16 A. No. I believe this is the only report 17 dealing with dated core data. If that is a correct 18 assumption, then the answer is in fact no. 19 Q. Well, when you say "dated core data," do 20 you mean data that's been collected and dated? I'm 21 not sure? 22 A. Yes. The data that have been collected 23 and dated in the cesium 137 method, I believe are 24 only used in the report we just discussed. You have 25 a report, nondated core stuff in February 26th and 340 1 May 5th reports -- 2 Q. Is that marked as an exhibit, I believe? 3 A. I don't know if you have discussed those 4 yet. 5 Q. Look at your left hand, the top -- 6 A. No, this is a direct quote out of what's 7 in here. 8 Q. Where are you reading from? 9 A. The top of this. All I'm trying to 10 determine is there's no dated core analysis in any of 11 the -- I am reading from Exhibit 6 -- that there's no 12 dated core data discussed in any of the previous 13 reports. I believe that's the case, but you can 14 confirm that by having me take a quick look at those. 15 (Pause) 16 A. It appears from the title, it's all 10 17 centimeter data which is fixed depth; February 26th 18 I'm sure as well. That appears to be a correct 19 statement, that the dated core analysis is solely in 20 the report we just discussed. 21 Q. You conducted a different analysis of the 22 phosphorus in WCA-2A, right? 23 A. This is all 2A. 24 Q. No, I know that, but you said dated core 25 analysis. What other analysis did you do in WCA-2A? 341 1 A. The analysis we discussed earlier 2 pertained to fixed depth data with Reddy's, which was 3 10 centimeter. 4 Q. We were talking about Exhibit 6 which was 5 the dated core data analysis that was done. 6 You had some unresolved issues and I will 7 let you look at them. 8 A. Yes. 9 Q. Did you resolve them? 10 A. They are in resolution now. All of those 11 issues have to do not with our analysis, but with the 12 data. 13 Q. But it would affect the analysis, would it 14 not? 15 A. Yes. 16 Q. Now, if the bulk density which was 17 measured or calculated turned out to be inconsistent 18 between Reddy and the Richardson sample sets, what 19 affect would that have on your analysis? 20 A. We will probably redo it. I think you can 21 expect that analysis will be redone. 22 Q. This whole analysis, we are talking about 23 Exhibit 6? 24 A. Any of the numerical results, maps and so 25 on, they will probably be redone there for exactly 342 1 that reason. 2 Q. You expect to redo it because you believe 3 the data that you get back will be inconsistent? 4 A. We have identified the inconsistency 5 there. 6 Q. You have identified the inconsistency? 7 A. Yes. 8 Q. What inconsistencies did you find? 9 A. I believe they are listed, 1, 2, 3, 4, 5, 10 or only 4? 11 Q. In No. 2 where you state that all reported 12 replicates in the Richardson sample set appeared to 13 be smaller in magnitude than that of the original 14 sample with which they are associated? 15 A. That's an issue. These are all data 16 issues. 17 Q. That's what I thought. 18 A. There's an inconsistency in the bulk 19 density that I think is identified that's a concern. 20 Q. It has been identified? 21 A. I think it is identified there as one of 22 those. 23 Q. Okay. You stated this inconsistency would 24 affect estimation of phosphorus mass and the 25 deposition rate? 343 1 A. Yes. It is not entirely clear at this 2 point whether it affects the mass or whether it 3 affects the concentration actually. 4 Q. So you plan to be working on it now, then? 5 A. There have been inquiries made of 6 Richardson and Reddy. 7 Q. Did you make the inquiry? 8 A. No, John Davis's office made the 9 inquiries. 10 Q. What are you trying to find out, how the 11 measure was calculated? 12 A. Simply an issue of why they get such 13 inconsistent bulk density estimates. 14 Q. These are measurements in the same 15 location? 16 A. Some of them were at very close to the 17 same location, yes. 18 Q. You figure that should be pretty close? 19 A. We figure they should be closer than they 20 are. 21 Q. You are trying to find out maybe if you 22 measured them incorrectly? 23 A. Or there was something about the way in 24 which they were reported which was not correct or did 25 not characterize properly our interpretation of what 344 1 they are saying is not correct or some problem. 2 Q. Were they being vague? It seems they can 3 tell you how they did it, they can both tell you how 4 they did it and you can make a decision whether or 5 not it was done consistently or not? 6 A. Well, the way these things are largely 7 handled in the absence of the legal process is that 8 one simply gets on the phone and calls and asks and 9 says, Look, all Richardson's numbers tend to be 10 higher or tend to be lower than all of yours, do you 11 have some good idea why? 12 And you usually get some response back, 13 Yes, we used such and such method and you used some 14 other method, and that sort of thing. 15 Q. It didn't happen here? 16 A. It doesn't happen quite so easily in this 17 context, but the inquiries have been made. 18 Q. How long would it take you to complete 19 your analysis after you have received the 20 information? 21 A. A few days, it won't take long. 22 MR. FITZGERALD: As with the possible 23 future reliance of the witness on the Burns & 24 McDonnell reports and the Reddy February '94 reports, 25 should the witness in fact engage in the additional 345 1 analysis, we would ask to be advised of that 2 immediately so we can arrange a mutually agreeable, 3 if not convenient, time to pursue the alterations in 4 whatever opinions or views this work might cause. 5 MR. LOREDO: I obviously concur with 6 counsel. 7 BY MR. LOREDO: 8 Q. You pointed out in issue No. 2 about the 9 replicates that appear to be smaller in magnitude 10 than the original sample -- you bring that up as an 11 issue. 12 What would that affect, or can you 13 elaborate on this issue that you set forth in your 14 analysis, in your report? 15 A. It is a quality control issue. It is 16 simply the kind of thing when you go through somebody 17 else's data and you see something that doesn't look 18 quite right, instead of just grinding the numbers 19 into the program, as students are sometimes want to 20 do, one raises the issue and tries to find out what's 21 going on there and if there's some explanation, 22 things of that nature, that something looks funny 23 there. 24 Q. A replicate, is that just a redoing? 25 A. Yes. I don't recall exactly how they did 346 1 their replicates, I believe they essentially split 2 their samples physically and took one part and 3 analyzed it and took the other part and analyzed it. 4 Q. Issue 3 speaks for itself. You are 5 looking for two additional stations that were 6 collected but were not available for the analysis of. 7 You are waiting for that data also? 8 A. Yes. 9 Q. The last one, the assumption that the core 10 data was collected at the same time period, have you 11 verified that assumption? 12 A. No. I don't think that's nearly as major 13 an issue as the other two, though. 14 Q. When you say "other two," do you mean -- 15 A. Or three issues, excuse me. 16 Q. I will go to the report which you have 17 marked No. 5 and we will mark it Exhibit No. 8, and 18 it is titled: Integration of Richardson and Patrick 19 Data Into Spacial Analysis of WCA-2A and WCA-3 10 CM 20 Phosphorus Core Data. 21 (Lettenmaier Exhibit 8 was marked for 22 identification) 23 BY MR. LOREDO: 24 Q. Doctor, take a minute and look at that and 25 give me a synopsis of the conclusions within the 347 1 report and we will go from there. 2 (Pause) 3 A. That simply augments the earlier analysis 4 with some additional samples taken from other 5 sources, and also extends the analysis to a small 6 part of WCA-3. 7 Q. It augments which analysis? 8 A. The previous analyses regarding 10 9 centimeter core data which are in reports we have not 10 yet discussed, and the first one you discussed. 11 Q. Was the intention of the work to develop 12 maps again of what phosphorus in WCA-2 and now also 13 WCA-3? 14 A. That's correct. 15 Q. Were you using GEOEAS here? 16 A. Yes. 17 Q. If I recollect correctly, for WCA-3, was 18 there a lot of data available? 19 A. No, I think it indicates there that the 20 amount of data that was available there was 21 relatively small. 22 Q. So the kriging work that was done there -- 23 I guess we are really estimating now for WCA-3? 24 A. For WCA-3, there does not appear to be a 25 lot of spacial structure based on the observations. 348 1 Q. When you say "spacial structure," what do 2 you mean? 3 A. Basically spacial trends. 4 Q. You can't determine the trend? 5 A. In the 2A there's a clear variation in 6 phosphorus concentrations as you go away from the 10 7 structures and go further south. That sort of 8 dependence is not nearly as set in the small part of 9 WCA-3 for which there were data. 10 Q. But in WCA-2, there is a trend from the 11 structures into I guess interiors? 12 A. Yes, our analysis and all the other 13 analysis shows that. 14 Q. So there is in fact a relationship -- 15 well, is there a relationship from the structures 16 into the interior areas of WCA-2? 17 A. Yes. 18 Q. What is that relationship? 19 A. You have maps there that show that. 20 Q. Well, I will give you the maps and can you 21 tell us on the record. 22 A. Well, this is WCA-3 that you just showed 23 me. 24 Q. Oh, I didn't -- 25 A. Do you want the ones for 2? 349 1 Q. Yes, I'm sorry. 2 A. Well, figure 8 is the figure that has the 3 kriging estimates of total phosphorus standard 4 deviation, and if you look at that, it shows what the 5 high values tend to be up here and the low values are 6 down here. 7 Q. The map you have shown me is the kriging 8 estimate of total phosphorus. Is there another one 9 done for standard deviation? I turn the page and 10 figure 9 does that, I assume? 11 A. I have to look at figure 9. 12 Q. You showed me figure 8. I thought that 13 was the kriging analysis total phosphorus 14 concentration? 15 A. Yes, I thought that's what you wanted. 16 Q. But you had stated that was the standard 17 deviation? 18 A. No, I said this was the estimate here, 19 figure 8, and that the values tended -- are higher at 20 the top and lower at the bottom. 21 Q. Okay. 22 A. Then you asked what the spacial patterns 23 were. 24 Q. Okay, and figure 9, is that the standard 25 deviation? 350 1 A. Figure 9 is the kriging standard deviation 2 that accompanies figure 8. 3 Q. That's interesting, the standard deviation 4 seems to be pretty close throughout WCA-2A? 5 A. Pretty close? 6 Q. I'm sorry, it looks like the range is 7 between 209 -- between 290 and 330 total milligrams 8 per kilograms of total phosphorus? It seems like the 9 deviation throughout WC/2A is pretty much the same. 10 A. That depends on where you are located. I 11 think you are looking at the particular labels that 12 the programmers put on contours which may be a little 13 bit misleading. 14 In fact, the kriging standard deviation is 15 zero at the exact points that the samples were 16 collected and becomes larger as you go away from 17 that, and I can't tell from some of these -- unlabled 18 for some reason -- they tend to be fairly close in 19 the area you are looking at, is that that's a 20 relatively uniform grid of samples through that, if 21 you look at where the samples were actually 22 collected. 23 Q. Again on figure 9, assuming a normal 24 distribution? 25 A. There's no assumption of distribution in 351 1 there. The only assumption of distribution comes in 2 if you want to start talking about probabilities. 3 Q. But when we talk about standard deviation, 4 aren't we also talking about probability? 5 A. No. 6 Q. We are not? 7 A. Not unless you ask to interpret that. 8 Q. If I asked you to interpret the standard, 9 the kriging standard deviation, then you would 10 determine probabilities for it? 11 A. I wouldn't -- if you asked what the 12 probabilities were, then we need to talk about 13 distribution. 14 Q. Why in figure 10 did you give us a 15 plotting or graph of total phosphorus concentration 16 plus one half kriging standard deviation? 17 A. Oh, that was simply -- that's associated 18 with some other reports you haven't talked about yet 19 where that was also done. The idea was to determine 20 the area affected or the area within a certain 21 contour and to get some idea of what the variation in 22 that area might be as reflected by the uncertainty in 23 kriging estimate. 24 Q. What does it mean to you, what does figure 25 10 represent then? 352 1 A. It represents exactly what it says. It is 2 the map of all the numbers in figure 8 plus half of 3 all the numbers in figure 9. I guess it is plus, I 4 thought it was minus actually. 5 Q. Figure 11 is, your next figure, total 6 phosphorus concentration less one kriging standard 7 deviation. 8 A. That's mine one standard deviation in 11 9 and plus one half in figure 10. 10 Q. You did this due to a prior analysis? 11 A. Well, the motivation is consistent with 12 the earlier analysis. All is done is that we 13 re-Kriged the earlier analysis using, in addition to 14 Reddy's data, the Richardson and Patrick data. 15 Q. So if I look at figure 10 versus figure 16 11, the northeast corner point, you are telling me 17 phosphorus, going from the plus half standard 18 deviation to a minus one, can be anywhere from 1200 19 to 1600 total phosphorus in milligrams per kilograms? 20 A. I am not sure what you are looking at. 21 Q. I am looking at the northeast corners of 22 both figure 10 and 11. 23 (Pause) 24 A. Northeast, I see a 1400 line, a 1600 line, 25 perhaps something like that; and -- 353 1 Q. An 1100 line and a 1200 line? 2 A. I see lines that are more like 900, a 3 thousand, yes. 4 Q. That would be a range then? 5 A. Yes. If you went and took the same point 6 and took the values that were used to generate those 7 contour plots, you would get the range between plus 8 one, minus one standard deviation and plus one half. 9 Q. Have you tried doing it plus or minus two? 10 A. We had considered that. When you subtract 11 two standard deviations, you end up with negative 12 concentrations and some other peculiarities, so it 13 doesn't make much sense. 14 Q. Did you do that calculation? 15 A. One can eyeball that. You don't have to 16 go through numerically and draw the maps to see that 17 there will be a problem with that. We considered 18 that, we considered using bigger values and decided 19 not to. 20 Q. When you eyeball, I guess taking the total 21 phosphorus -- obviously my eyeballing -- when you say 22 "eyeballing," would you like eyeball figure 8? 23 A. I go and look at the kriging standard 24 deviations there and I see a lot of numbers like 300 25 in a particular point, and then I go over and look at 354 1 figure 8 and I see a number like 400 which is the 2 best estimate. 3 And if I took two standard deviations, 4 which would be like 600, and subtract from 400, I 5 will have a negative number which won't make a lot of 6 sense so we didn't do that. 7 Q. You picked 400 in figure 8, that's 8 basically the middle? 9 A. I am just giving you an example. I am 10 just looking somewhere there, and I see some 400s, 11 and if I go over to the other plot at the same 12 location where I see a 400 and say it is something 13 like 300, and double that and subtract it, I realize 14 I would be in trouble. 15 Q. Moving you to the report which you have 16 had marked No. 4 on top and we will make it Exhibit 17 9, titled: Estimation of Soil Phosphorus and 18 Topography in the WCAs. 19 (Lettenmaier Exhibit 9 was marked for 20 identification) 21 BY MR. LOREDO: 22 Q. It says it is drafted May 5th, and it says 23 "corrected 6-7-93." Is that your handwriting, 24 "corrected"? 25 A. That is in fact my handwriting, yes. 355 1 Q. Doctor, if I can give you this report and 2 if you can give me a synopsis of the purpose of it 3 and what you concluded or found in the review and 4 analysis? 5 A. Well, there are three parts to this 6 report. 7 Q. Start with Part 1. 8 A. Part 1 estimates the distribution of -- 9 apparently there's a February 26th report which we 10 haven't discussed yet, and it appears there's some 11 update on some of the analysis of the area that would 12 be enclosed within contours representing the kriging 13 estimate of either a fixed mass rate or a fixed 14 concentration based on the minus half standard 15 deviation or plus one half and minus one half. 16 My guess is that it is probably minus one 17 and plus one half, the same as the other report. 18 Q. Are we guessing? 19 A. Well, I would have to go look at the other 20 report, and I probably would have to go look at the 21 computer files to verify. 22 Q. When you say "other report," you mean the 23 report dated February 26th, '93? 24 A. Right, and the one you just went through. 25 Q. Do you want to take a quick look? 356 1 A. I probably need the computer files to 2 verify that. 3 Q. Your guess is -- 4 A. I was surprised that the other report said 5 that it was minus one and plus one half because I 6 believe that it was minus one half and plus one for 7 the reason that I indicated, to avoid having negative 8 concentrations. 9 And this report is saying plus one half 10 and minus one half, so I would have to verify it, 11 because these were done in a consistent manner, and I 12 think what was done in the report may not reflect the 13 analysis that was actually done, it may not be labled 14 correctly. 15 Q. Are you telling me the report we marked as 16 Exhibit 8 under the Richardson and Patrick data, 17 where figure 10 was plus one half and figure 11 was 18 minus one half, that needs to be verified because 19 that may not be correct either then? 20 A. I didn't think it said plus one half or 21 minus one -- I think it said plus one -- let me look 22 at that one. 23 Q. You have got me confused now. 24 A. Figure 10 says plus one half and figure 11 25 says minus one. This one here says minus one half 357 1 and plus one half which leads me to believe that 2 there's some inconsistency in labeling. 3 Q. Exhibit 8 is the latter report, right, 4 dated June 7th, 1993? 5 A. I'm not sure -- this is Exhibit 9, I'm not 6 sure what Exhibit 8 is. 7 Q. Exhibit 9 is dated when? 8 A. Exhibit 9 is dated May 5th, '93. 9 Q. And Exhibit 9 is saying minus one half and 10 plus one? 11 A. The exhibit you have, Exhibit 8, says plus 12 one half and minus one. 13 Q. Right, and Exhibit 9 is the reverse? 14 A. Well, I believe it is simply an 15 inconsistency in the text. Figures 2-6 and 2-7 say 16 plus one half and minus one, which is consistent with 17 what you have in the other report. There's something 18 in the text here that says -- 19 Q. What page are you reading from? 20 A. If you look at figures 2-6. 21 Q. I looked at them. 22 A. Fine. Those are consistent with what we 23 just discussed from the other report. 24 Q. Right. But within the text -- 25 A. Within the text here I see down at the 358 1 bottom of page 2, the paragraph that's labeled 2.0, 2 it says: Plus one half kriging standard deviation 3 and minus one half kriging standard deviation. 4 It appears that should say plus one half 5 and minus one, which would be consistent with the 6 other report. That probably needs to be verified. 7 Q. I see what you are saying. 8 A. What's labeled on the figures is 9 consistent with the other report, what it says in the 10 text is not. I suspect what's labeled in the figures 11 is correct. 12 So based on those -- where were we -- you 13 asked what this does relative to the other report, 14 and I have to look at the other -- 15 Q. We start with Part 1, and you said it was 16 an estimate of distributions, and then you referred 17 to a prior analysis, and I think the prior analysis 18 is the one dated February 26th, '93. 19 A. I have to look at the February 26th report 20 to see if there are any area estimates made in that 21 report. 22 Q. Why don't we do this then. 23 Why doesn't the court reporter mark as 24 Exhibit 10, the report titled: Estimation of Soil 25 Phosphorus in WCA-2A, and that's the report dated 359 1 February 26th, and maybe we will let you do this 2 report first. 3 A. Either that or if you can just hand that 4 to me for a minute, I can tell you. 5 (Pause) 6 A. The earlier report in fact did do the plus 7 one half and minus one standard deviation analysis. 8 Q. Let the court reporter mark that first. 9 (Lettenmaier Exhibit 10 was marked for 10 identification) 11 BY MR. LOREDO: 12 Q. Doctor, what might not be a bad idea is to 13 let you -- either way you would like to go, we have 14 marked as Exhibit 10 the earlier report -- 15 A. You can either start with 10 or go with 9, 16 it doesn't make any difference to me. 17 Q. Excuse me? 18 A. Either 9 or 10, you can do it either way. 19 Q. I will let you pick, whichever is easier 20 for you. Because it seems that what we have marked 21 as Exhibit 9 is an update of the Exhibit 10 report. 22 A. Right. Well, I would be just as happy if 23 you went through 9 now that I see what's in here. 24 Q. Go ahead. 25 A. Figure 9 basically goes through -- the 360 1 first part has to do with estimating the range of 2 area that's within specified values of either 3 phosphorus mass, which that critical value is taken 4 as four metric tons per square kilometer, or 5 concentration, alternatively, which the critical 6 value is 600 milligrams per kilogram. 7 And those are, more or less, taken from 8 reports of Richardson, some analysis he had done and 9 some discussions with him as to reasonable values of 10 an affected area, of an area that would be affected 11 by the 10 structures. 12 We estimated in this report for several 13 variations and several different assumptions in the 14 kriging and then also for the plus one half and minus 15 one standard deviation to get some estimate of how 16 variable those areas might be, and in table 2 there, 17 we characterize how, essentially, how the area would 18 change based on the area within the fixed contours. 19 For instance, the area based on the 20 phosphorus mass, the four metric tons per square 21 kilometer, based on plus one half of a standard 22 deviation, it would be 187 percent of the best 23 kriging estimate, and the minus one standard 24 deviation area enclosed would be 37 percent. It was 25 a rather wide variability in those estimates. 361 1 Q. When you say 187 percent of the acreage -- 2 A. Yes. 3 Q. If the acreage was 18,520, are you saying 4 the acreage should then -- 5 A. If one goes to the map, the kriging map, 6 and simply goes and draws a line around the four 7 metric ton per square kilometer contours and shades 8 all of that area for the kriging estimate and finds 9 one value of the acreage, how many acres are within 10 that, and now does the same thing for the kriging 11 estimate plus one half of standard deviation, we are 12 saying the No. is 187 percent or 1.87 times the 13 earlier one, the same way with the minus one standard 14 deviation. 15 That sort of analysis is done when one 16 also uses concentration, and you find there it looks 17 like in this one particular case, 176 percent and 55 18 percent when you look at the 600 micrograms per 19 kilogram contour as opposed to the four metric ton 20 per square kilometer. 21 Q. These are all used in anisotropic kriging? 22 A. Well, in table 1, we compared various 23 kriging estimates, and this is the best estimate 24 without the plus standard deviation, minus standard 25 deviation and so on; within the either four metric 362 1 tons per square kilometer of 600 microgram per 2 kilogram contour, that was done for isotropic and 3 anisotropic methods. 4 Over in table 2, we did the comparisons 5 with the plus one half and minus one only for the 6 anisotropic which we believe is a better method. 7 MR. LOREDO: Let's take a break. 8 (Recess) 9 BY MR. LOREDO: 10 Q. We were talking about table 1 and 2 in 11 Exhibit 9 and you were telling us about the area of 12 the table 1 and then the plus one half of kriging 13 standard deviation -- I'm sorry -- plus one half and 14 then less one kriging standard deviation. 15 Now, did you do anything with Reddy's 16 contours of concentration? I notice in table 1, the 17 fifth item? 18 A. Well -- 19 Q. Can you do a kriging on that? 20 A. Wait a minute, you have to look for a 21 description here. Reddy's Kriging Estimate Digitized 22 by ESP. 23 Q. That's the 24,000 acres plus? 24 A. Yes, they determined that by digitizing 25 the map out of his report, that one confining the 363 1 area within the 600 micrograms per kilogram, that it 2 is 24,762, simply reported there for comparison. 3 Q. You said you digitized the report? 4 A. I didn't say I digitized it, it says ESP 5 did it. 6 Q. What are the possibilities of errors being 7 in a digitized report? 8 A. You would have to ask them. 9 Q. Is it possible to have errors? 10 A. What do you mean by "errors," of what 11 magnitude? 12 Q. Well, is it possible to have errors? 13 A. There's errors in everything in this world 14 virtually. If you asked me how large are the errors 15 likely to be in my best professional judgment of what 16 they would get by digitizing the contours, I would 17 say it is unlikely they are much larger -- they are 18 about five percent, probably not larger than about 19 five percent and, quite likely, smaller than that. 20 Q. Where do you get the Bathymetric data for 21 the kriging effort? 22 A. We are moving on to that? Those are 23 provided. 24 Q. Provided by who? 25 A. By ESP. 364 1 Q. What gives a better estimate of the total 2 phosphorus in the WCA-2, using the anisotropic 3 diagram or the isotropic? 4 A. I believe the anisotropic variogram is 5 more appropriate. 6 Q. That takes into account the angles then? 7 A. Yes. 8 Q. And this associates from the linearity of 9 the data? 10 A. There is evidence that the variogram 11 depends on direction, and there's some analysis in 12 that report and in some of the other ones that 13 substantiates that. 14 Q. That there is? 15 A. There is, there tends to be more 16 dependence in the general direction of flow than 17 there is lateral to that direction of flow, as one 18 would expect on physical grounds as well. 19 Q. If there's more dependence then -- when we 20 say "more dependence," are we saying that there is a 21 relationship, I guess, in the southeast, everything 22 going south? 23 A. The variogram is different generally in 24 the north-south than it actually turns out to be -- I 25 can't remember exactly -- slightly angled off from 365 1 the north-south, but generally north-south 2 directions, the variogram is different and quite 3 substantially so than it is in the general east-west 4 direction. 5 Q. Figure 2-6 and 2-7? 6 A. Yes. 7 Q. Here you are saying Estimated Distribution 8 of Total Phosphorus -- kriging plus one half kriging 9 standard deviation? 10 A. Right. 11 Q. Where is or what is the estimated 12 distribution of total phosphorus without the addition 13 of the one half kriging standard deviation? 14 A. I'm sorry, say that again. 15 Q. Maybe I'm lost. 16 In most of the reports we started with a 17 kriging estimate, and then we did another one with 18 the standard deviation of plus or half and then we 19 did another one minus one? 20 A. Yes. 21 Q. Is that it? It is figure 2-4 -- I guess 22 what I am getting at, Doctor, is when I look at 23 figure 2-6 -- I am trying to backtrack here -- what 24 figure, what diagram you have in your report that you 25 extend on these deviations. 366 1 Where is the beginning product, where is 2 the estimation of total phosphorus, the estimated 3 distribution of total phosphorus mass in WCA-2A -- 4 MR. BLANK: Do you know what he is asking? 5 THE WITNESS: Yes, I know. You are asking 6 is there a map in here for the total phosphorus mass 7 without the plus one half or without the minus one? 8 BY MR. LOREDO: 9 Q. Yes. 10 A. It doesn't appear there is, but that 11 certainly is in the computer, and I am quite sure you 12 have the files with it. We clearly deemed it not to 13 be necessary to produce every map; we would have a 14 mountain of paper. 15 Q. I understand that, but for my edification -- 16 A. To go krig that from this if one wants -- 17 I guess you can't because you don't have the standard 18 deviation here. We didn't see that it was necessary 19 to go krig that. 20 Q. You are saying you provided that. Is that 21 on the UNIX or on the disks? 22 A. On the disks, all the kriging stuff is on 23 the disks. 24 Q. You would find it on there? 25 A. You would find it on there. 367 1 Q. I had a pattern going and you messed up 2 the whole thing. 3 A. I don't know exactly why we didn't -- 4 probably because I went through and said, Look, this 5 goes on long enough already, that's not one we need. 6 Q. Let's go to Part 2. 7 A. We are already there. 8 Q. I guess that starts on page 15 titled: 9 Kriged Bathymetry and Sediment Thickness for WCA-2A 10 and S 9 Area? 11 A. Yes. 12 Q. Can you tell us what we are doing here? 13 A. Yes, we are developing a map of the 14 sediment thickness and the top of sediment depth in 15 WCA-2A and the S 9 area of WCA-3. 16 Q. The sediment thickness and the top of the 17 sediment depth? 18 A. The top of the sediment elevation, I'm 19 sorry. 20 Q. What do your results show? 21 A. We developed a map of those on a 22 particular grid. I'm not sure what plots are 23 included in here. There are a lot of intermediate 24 results. 25 The bottom line is figure 3-6, Estimated 368 1 Distribution of Sediment Surface Elevation in 2A, and 2 3-12, Estimated Distribution of Sediment Thickness in 3 2A, and then their equivalent plots for for S 9, 3-14 4 and 3-17. Those are the results of that analysis. 5 Q. Let's go back to figure 3-6. 6 A. Right. 7 Q. Now are when we say "sediment surface 8 elevation," what are we really talking about? I 9 don't understand. 10 A. That's the elevation of the top of the 11 sediment. I mean, if you go out there and it is at a 12 time of the year when things are wet and you put out 13 a little stick and you push it down until it first 14 intercepts the top of the sediment, that's the top of 15 the sediment elevation. 16 Q. And you are measuring which way, how far 17 your stick goes in? 18 A. I believe this is all based on a survey, 19 so that is -- those numbers are above some data which 20 is going to be mean sea level probably. 21 Q. So this is a measurement from basically 22 where you are touching the ground, the bottom? 23 A. If the sediment weren't swishy and it were 24 a concrete surface, that is the elevation of that 25 surface. 369 1 Q. Now since it is swishy? 2 A. Swishy, somebody has gone out there and 3 determined just where the top of the sediment is. 4 Q. In your analysis, these numbers, 10, 11.6 5 -- centimeters? 6 A. I believe those are feet, they are from 7 some datum which was specified probably above mean 8 sea level. 9 Q. Whose data, is this ESP data? 10 A. ESP processed it. There was a contract 11 let to collect the data. You can probably ask 12 counsel who performed the actual survey. 13 Q. I'm sure counsel doesn't want to be 14 deposed here. 15 Is it in your report? 16 A. You would have to look. The data were 17 provided to us by ESP. It is certainly something I 18 can find out for you, and if you ask Mr. Blank, I'm 19 sure he could tell you right now who the contractor 20 was that did the survey. 21 Q. Then you pointed me to figure 3-12, I 22 believe? 23 A. That is correct. 24 Q. Sediment thickness is what, the mushy 25 part? 370 1 A. That's the depth of the sediment down to 2 bedrock, also as provided by the contractor. I 3 believe those depths are feet. 4 Q. The question that pops in my mind is why 5 or what does this represent to us, what's the 6 significance of this, if any? Or do you know? 7 A. Do I know? 8 Q. Yes. 9 A. Why anyone would be interested in how much 10 sediment is out there? 11 Q. Or sediment thickness. 12 A. Well, one can only imagine that somebody 13 might be interested in the total amount of phosphorus 14 that was out there, for instance. 15 Q. And that affects it, the sediment 16 thickness? 17 A. Presumably there is not a lot of 18 phosphorus in the bedrock, at least that's not very 19 mobile, so knowing how much sediment out there might 20 be of some importance. One might also be interested 21 in knowing something about water depth. 22 Q. I thought at least some of the analysis of 23 Dr. Millard was that there was no relationship 24 between the water depth and phosphorus? 25 A. You took the deposition of Dr. Millard, 371 1 I'm not sure what particular aspect of the analysis 2 he would be referring to. 3 Q. You believe there is a relationship 4 between the water depth and phosphorus? 5 A. I haven't done an analysis. 6 Q. Does your report consider water depth? 7 A. No. 8 Q. Figure 3-15. 9 A. Yes. 10 Q. That's showing me the location of the S 9 11 sampling location? 12 A. That's correct. 13 Q. It is an useful map. 14 Figure 3-17? 15 A. Yes. 16 Q. Estimated Distribution of the Sediment 17 thickness in S 9 Area. I guess from this, we are 18 looking at thickness between 2.2 and 2.6, and I'm not 19 sure in what measurement you are looking at again. 20 A. Feet, I believe. 21 Q. When you say "estimated distribution"? 22 A. Spacial distribution, it is a map. 23 Q. It is by space? 24 A. You could have retitled that map. 25 Q. This is not kriged then? 372 1 A. It is kriged. 2 Q. How many samples sites on the S 9 area? 3 A. I would have to go over here and count up 4 the number in figure 3-15. 5 Q. 3-15? 6 A. No, in figure 3-15, that shows where they 7 are exactly how they are labeled. I don't recall. 8 It appears there's something on the record 9 of 30 or 40 of them. It probably says if you look in 10 here closely enough. 11 (Pause) 12 A. The specific number is in the computer 13 files, and I don't recall it. Given a minute or two 14 in looking at figures 3-16 and the separation 15 distance, shows me that probably a separation 16 distance of somewhere in the 300 meter range. Then I 17 have to go back and figure out what the spacial 18 extent of the S 9 area is. 19 Q. It is a very small area that was sampled? 20 A. A relatively small area, yes. It was 21 relatively a fine mesh of samples in the area, and 22 approximately 30 to 40 is a good place to start. We 23 could find the exact number for you if it is 24 critical. 25 Q. If I remember correctly, you said it was 373 1 at least 30 when you are going to do a kriging? 2 A. Yes, I think I said there appears to be 30 3 to 40. 4 Q. I know that. 5 You can at least say there's probably at 6 least 30? 7 A. If it was substantially less than 30, we 8 wouldn't have attempted a kriging. 9 Q. What would Part 3 of this report be? 10 A. We have just gone through Part 3, the 11 third part being section four. 12 This is just the kriging of the 13 Loxahatchee soil phosphorus data. The method is 14 similar to what was done in WCA-2, 10 centimeter 15 data. 16 Q. What page is that? 17 A. It starts on 32. 18 The method is basically the same. It went 19 through and estimated anisotropic variograms, used 20 GEOEAS to go construct the map of the 10 centimeter 21 phosphorus mass concentrations, and figure 4-5 says 22 kriging error, which is what people sometimes call 23 it; kriging standard deviation would be somewhat more 24 consistent. 25 Q. Is this plus one -- 374 1 A. No, there's no plus one or minus half or 2 any of that. We were not interested in the area 3 enclosed, we were simply interested in mapping. 4 Q. What are the units? I am looking at 5 figure 4. 6 A. Figure 4, mass is going to be metric tons 7 per square kilometer. I believe that's consistent. 8 Masses are in metric ton per square kilometers and 9 concentrations are in micrograms per kilogram. 10 Q. I didn't understand the explanation when 11 we talked about the error. You said we weren't 12 looking at spacial -- take me back there again. I 13 was asking why didn't you again do plus one half 14 minus one standard deviation? 15 A. Because the analysis of the enclosed area 16 isn't particularly relevant or hasn't come up in that 17 area. 18 Q. I'm kind of lost. You said it didn't 19 matter to do the standard deviation. Take me back 20 over that again. 21 A. Well, the whole motivation for doing the 22 plus half and minus one standard deviation was 23 looking at an area that was enclosed within a 24 particular contour. 25 That was done in 2A because there's been a 375 1 concern about the affected area and where the 2 phosphorus that's entering the area from the 10 3 structures was falling out because of work that's 4 been done there associated with settling rate. 5 That's not a concern in the Loxahatchee, so we didn't 6 perform that analysis. 7 Q. When I look at figure 4-5 -- I am looking 8 for a number, I think I find 3.4? 9 A. Yes. 10 Q. What does it stand for? 11 A. Metric tons. 12 Q. It is 3.4 metric tons? 13 A. That's the one standard deviation. 14 Q. That's a one standard deviation? 15 A. That's the kriging standard deviation. 16 Q. So it would be a plus one? 17 A. No. The kriging -- remember, as I 18 explained, kriging gives you an estimate of the value 19 at any arbitrary point. It also gives you the 20 standard deviation of that estimate. So you get or 21 can make two separate contour maps, one of the 22 kriging estimates, the other the standard deviation. 23 4-5 is just a standard deviation. 24 Q. I'm a little stuck, Doctor, at least when 25 we talked about it before -- and I guess maybe I 376 1 haven't conceptualized the difference in WCA-2 and 2 WCA-1, but when I look at the kriging error, the 3 errors that were developed for deviation in figure 4 4-5, we were talking about 3.4, and if I go back to 5 figure 4-4 and I either go plus or minus the one, I 6 end up with a negative, and I thought that's 7 something you wouldn't want to do. 8 A. Well, why would one go plus or minus? 9 Q. I thought that's what deviation did -- 10 A. No. 11 Q. You don't -- 12 A. If one wants to make interpretations of 13 confidence bounds, then one needs a distribution. 14 Standard deviation says nothing about distribution. 15 You can have a standard deviation of extreme value 16 distributions, of many other distributions which are 17 assymetric. 18 Q. So the standard error created by the 19 kriging -- 20 A. Standard deviation. 21 Q. Standard deviation -- instead of using the 22 word error -- is that always plus? 23 A. No, there's a mathematical definition of 24 what the kriging standard deviation is. The square 25 root of the variance, and the variance is the 377 1 expectation of the estimate minus its mean squared. 2 That's a mathematical definition. If one 3 has a probability distribution, one can then go make 4 some interpretations, some probabilistic 5 interpretation that would allow you to make perhaps 6 statements about confidence intervals. 7 But the fact that minus one standard 8 deviation is negative, doesn't say anything other 9 than that the distribution must be positively skewed, 10 or at least most likely a mathematician could 11 probably come up with a percent, with a pathological 12 case where that didn't imply a positive skewness, but 13 generally it would. We engineers try to avoid 14 dealing with pathologicals. 15 Q. Did you attempt to use the kriging error 16 for total phosphorus mass in WCA-1 using the 17 isotropic variogram? 18 A. Is there a plot there? 19 Q. First, did you do it? 20 A. It falls out of the analysis, sure. We 21 did the analysis -- did we do an isotropic analysis, 22 is that the question? 23 Q. Yes. 24 A. I'm not sure whether we did or not because 25 there was strong evidence of anisotrophy. The main 378 1 reason we did it in the other cases is others who had 2 analyzed the data using an isotropic assumption. 3 To my knowledge, there are not other 4 analyses, or at least there are not other analyses, 5 kriging analyses of the WCA-1 data around, so we did 6 it the way we felt was best. 7 Q. I thought you said you did it all the way, 8 it is just that maybe in the report you just decided 9 to use or put in the anisotropic variogram? 10 A. Without going through my computer files, I 11 can't tell you whether there was ever an isotropic 12 analysis one for 1, there might have been. 13 Q. It wouldn't automatically be run? 14 A. It wouldn't automatically be run, no. 15 Q. And you didn't want to do it because it 16 would just be a waste of time? 17 A. Well, if the data showed that the 18 variogram is anisotropic, why would you want to go do 19 the analysis of an isotropic one? 20 Q. With respect to the variograms, did you 21 make any account for the different inflow structures 22 being in two different areas and possibly come up 23 with two different maps, contours, or you just let 24 them overlap? Or does my question make any sense? 25 A. The latter I think is correct. 379 1 Q. My question makes no sense? 2 A. It makes no sense to me. Perhaps you 3 could explain it a little more. 4 Q. In WCA-1, there's two major inflow 5 structures. 6 A. In WCA-1? 7 Q. Yes. 8 A. Yes, that's correct. 9 Q. In doing the isotropic variogram analysis, 10 would it be appropriate to first just do it under, 11 assuming there's just one structure and then assuming 12 there's another structure, and seeing what the 13 contours would be like separately? 14 MR. BLANK: Excuse me just a minute, I 15 thought I heard you say "isotropic"? 16 MR. LOREDO: I did say isotropic, maybe it 17 should be anisotropic. 18 MR. BLANK: I want to be clear what your 19 question is. 20 THE WITNESS: Who is doing the analysis 21 now? There's no assumption in the analysis about 22 where the structures are, where flow is coming from 23 or any of that. The anisotrophy is inferred from the 24 data. 25 There's a primary axis and a secondary 380 1 axis and one would expect the primary axis would more 2 or less line up with the direction of flow. 3 Q. You said there was a primary and 4 secondary? 5 A. Yes. 6 Q. What was the secondary? 7 A. Which is "orthogonal" to the primary. 8 Q. What does that mean? 9 A. 90 degrees. 10 Q. A right angle? 11 A. Yes. 12 Q. Did you take into account any other 13 degrees and angles, does it matter if it was 45 14 degrees? 15 A. The way things are set up is that there's 16 a primary and secondary axis, and if you are using an 17 anisotropic variogram, then you have one variogram 18 that applies to the primary axsis and another that 19 applies to the secondary axis, and the mathematics 20 take care of any angles that are in between. 21 Q. And the isotropic -- 22 A. Assumes they are all the same, so there's 23 only one variogram, it doesn't make any difference 24 what direction you have. 25 Q. So you have two sets of variograms in 381 1 anisotropic analysis or maybe several? 2 A. Yes, those are -- all these figures in 3 here pertain to the variograms in the different 4 directions. 5 MR. LOREDO: Let's take about five 6 minutes. 7 (Recess) 8 BY MR. LOREDO: 9 Q. Going back on the record. 10 Have we finished talking about that 11 report? 12 A. I'm finished, yes. 13 Q. That was Exhibit 9? 14 A. Right. 15 Q. How did that change from the report that 16 we have marked as Exhibit 10, what's the change? 17 A. What's in Exhibit 10? 18 Q. What's Exhibit 10 titled? 19 A. Exhibit 10 is titled Estimation of Soil 20 Phosphorus in WCA-2A. 21 Well, the only things that the previous 22 report pertains to that this one also pertains to are 23 the 2A concentrations. Most of what is in this 24 report is probably intermediate material which 25 introduces the idea of estimating the area within 382 1 specific contours within 2A, but that is then pursued 2 further in the subsequent reports. 3 So I think you are going to find for those 4 purposes, the subsequent report or reports are really 5 more relevant than this one. 6 Q. Did you change any opinions as to the 7 results of your analysis from that report? 8 A. I'm not sure that there's any opinion 9 stated in here. This is essentially a progress 10 report. We conclude in this report that things are 11 strongly anisotropic, and that had not been said 12 before because we hadn't done the analysis before. 13 This report is the one where we switched 14 to GEOEAS because the anisotropic analysis in GEOPACK 15 turned out to have bugs in it, and then this analysis 16 was carried on in the subsequent reports to get the 17 plus one or plus one half minus one analysis and so 18 on. 19 Q. How did you find out that GEOPACK had bugs 20 in it? ? 21 A. We ran it and it didn't work. 22 Q. The first time you ran it, it didn't work? 23 A. Well, if you are asking me specifically 24 where the program bombed and why it wouldn't work, I 25 don't remember the details. I remember the parts of 383 1 it that were advertised to do anisotropic analysis 2 would not work as advertised in the manuals. 3 Q. I'm trying to figure out when you found 4 out, when you started using it? 5 A. Before February 26th, 1993. We had 6 discussions with the fellow, who I think is down here 7 somewhere who developed the program, under contract 8 with EPA, and were given stories about his contract 9 hadn't been continued to finalize some work, and, 10 yes, he knew there were bugs in certain parts of it, 11 and we determined at that point probably that wasn't 12 the one to use. 13 Q. What was his name? 14 A. I don't remember. I think in one of those 15 reports there's a reference to the GEOPACK manual, 16 which is an EPA publication. 17 Q. Have you had any discussions with Dr. 18 McClave? 19 A. Yes. 20 Q. Pertaining to what? 21 A. I met with him several weeks ago. I think 22 I had one phone conversation prior to that. 23 Q. How many times have you met with Dr. 24 McClave? 25 A. Once. 384 1 Q. What was the nature of the meeting? 2 A. We discussed some work he had been doing 3 in trend analysis. 4 Q. What work is that? 5 A. He had been redoing some of the trend 6 analysis for the Park. 7 Q. Did you give him any suggestions? 8 A. I don't recall if I made any suggestions. 9 Q. Did he tell you what he opined to? 10 A. Well, I saw some of the results. 11 Q. What were they? 12 A. They were plots of some of the data and he 13 had some fits of curves to some of the data. 14 Q. This was strictly with the Park? 15 A. My recollection is that that's the case, 16 although I'm not certain of that. It is part of the 17 discussion I have the best recollection of. 18 Q. What do the plots tell you? 19 A. They seem to show in general the 20 concentrations had been increasing until the late 21 1980s and subsequently decreasing. 22 Q. Concentrations increase until 1980 and 23 subsequently thereto, have been decreasing? 24 A. Seem to be decreasing, yes. 25 MR. BLANK: Excuse me, counsel, did you 385 1 say 1980 or '80? 2 Q. 1980. 3 A. I'm sorry, late 1980s is what I said, 1989 4 approximately, between '85 and '89. 5 Q. Was there a theory on the turning point or 6 why there was a change that wasn't increasing any 7 more and it seemed to be decreasing? 8 A. No. 9 Q. Were there any other variables within his 10 analysis? 11 A. Than? 12 Q. Phosphorus and time are the only things 13 you seemed to talk about? 14 A. Other chemical constituents, factors? 15 Q. Yes. 16 A. I don't recall. 17 Q. What was your telephone conversation with 18 Mr. McClave about? 19 A. Just to determine the type of work he was 20 doing and that a meeting might be desirable. 21 Q. So you had the phone call first? 22 A. We had the phone call first and decided a 23 meeting in person would be a good idea. 24 Q. Where did you meet? 25 A. At ESP offices in Gainesville. 386 1 Q. This was several weeks ago? 2 A. Yes, it was very roughly around the first 3 of February. 4 Q. Of this month? 5 A. Of last month. 6 Q. We are in March now. 7 A. Yes. 8 Q. How long did you meet with him? 9 A. Several hours. 10 Q. Did he give you anything, such as 11 documentation, computer disks, notes? 12 A. I can't recall if I got a copy of his 13 plots or not. There may have been copies of plots. 14 I'm quite certain I didn't bring them back to Seattle 15 with me. 16 Q. What would you have done with them? 17 A. Review them possibly in my hotel room. 18 Q. And left them in the hotel room? 19 A. Yes. 20 Q. So you were able to analyze these plots 21 while you were in your hotel room? 22 A. Yes, a general picture of what he was 23 doing. 24 Q. How many plots? 25 A. He had lots of plots. 387 1 Q. Pages and pages of plots? 2 A. Pages and pages. 3 Q. You just left them in the hotel? 4 A. They were probably in the trash. 5 Q. They weren't any good? 6 A. No, I didn't say that. 7 MR. BLANK: Nice try, counsel. 8 Q. Was there any relationship to stage 9 analyzed? 10 A. That's what I don't recall. 11 Q. How about location of the phosphorus 12 concentration? 13 A. The stations? 14 Q. Stations -- I guess I may be confusing 15 this, but aren't there different stations within the 16 Park that you take samples from? And I am talking 17 about, is he saying at particular points in the Park, 18 the concentration has gone up and now it is coming 19 down? What area in the Everglades, is what I'm 20 looking for? 21 A. He had analyzed a number of stations in 22 the Park, most of which are -- possibly all the same 23 as the ones the Justice Department consultant has 24 analyzed. 25 Q. He has just taken raw data and just 388 1 plotted them? 2 A. No. He had fit some curves to them and 3 used some methods of analysis on them. 4 Q. What type of methods of analysis? 5 A. I don't recall precisely the method he has 6 used to fit the curves. 7 Q. Do you want to take a best guess? 8 A. A best guess? 9 Q. Of what you recall? 10 A. I don't think so. 11 Q. How about a first guess? 12 A. A first guess as to method? 13 Q. Yes. 14 A. He used probably some sort of least 15 squares analysis. 16 Q. Parametric analysis? 17 A. Yes. 18 Q. He did use the parametric analysis? 19 A. I believe so. 20 Q. What were the results of that, what did it 21 show? 22 A. Well, he had some curves, and the curves 23 went up and they went down. 24 Q. I thought you didn't agree with a 25 parametric analysis? 389 1 A. I don't necessarily. 2 Q. Did you tell that to Dr. McClave? 3 A. Well, there was some discussion of that. 4 Q. What did you tell him? 5 A. I don't recall exactly what I said, 6 probably to the effect of my preference would be for 7 a nonparametric analysis. 8 Q. Did he do a nonparametric analysis? 9 A. I can't recall if he had done a 10 nonparametric analysis or intended to, there was 11 discussion of that. 12 Q. What was his response to you when you told 13 him you didn't agree with the parametric analysis? 14 A. I don't recall. 15 Q. Anything I can give you to refresh your 16 memory? 17 A. Probably not; get him on the phone. 18 Q. Was there a log transformation? 19 A. He may have transformed the data, I don't 20 recall. 21 Q. Can you tell me what you do recall of your 22 meeting with Dr. McClave? 23 MR. BLANK: I think he already has, 24 counsel. Are you asking him again? 25 MR. LOREDO: I'm asking him again because 390 1 he seems not to recall too much of the meeting. 2 THE WITNESS: The only thing I recall is 3 there appeared to be something interesting going on 4 with the data, and it was probably worth doing some 5 additional analysis, and I think I indicated 6 yesterday what the nature of that analysis was that 7 he intended to do. 8 BY MR. LOREDO: 9 Q. Because it was a parametric analysis? 10 A. I'm sorry? 11 Q. What were the dangers? 12 A. the dangers of what? 13 Q. Of the analysis. You just said there was 14 a danger? 15 A. No, I didn't say anything, I didn't 16 mention the term "danger." 17 Q. Did Dr. McClave do a first order 18 regression? 19 A. First order regression? Can you define 20 what that means? 21 Q. You you want me to define what "first 22 order regression" means? 23 A. I have never heard of the term. 24 Q. You can't ask me questions. 25 A. I think you mean first order 391 1 autoregressive -- all one word. Somebody feeding -- 2 Q. Thank you for your help. 3 Did he do it? 4 A. I don't recall. 5 Q. We talked about the other day, you met 6 with Dr. Marin? 7 A. Yes. 8 Q. And you will review his work with respect 9 to the STAs? 10 A. Yes. 11 Q. I won't go over that. 12 Q. I have some closing things and I will let 13 Tom have fun. 14 I just want to confirm, and I know we went 15 over this, but just so I am clear, we have gone over 16 your reports, and I have an idea of the work you have 17 done, so I now would like you to respond to some -- 18 these are issues that your attorney has set forth 19 that you will testify to, and I just want to confirm 20 on what your testimony will be. 21 I am starting on page 17 of I think it is 22 the most recent pretrial disclosure of issues of 23 witnesses, and the first issue where your name is 24 listed as a possible witness -- you are at the head 25 of the list, and I would like you to explain what you 392 1 will testify as to this issue, the first one being 2 whether the SWIM Plan's phosphorus limiting standards 3 for the EPA are necessary to or will prevent 4 violations of state water quality standards, if any 5 are determined. 6 A. And you are asking me what my specific 7 testimony will be on that? 8 Q. On that issue, yes. 9 A. I'm not sure with it phrased that way 10 exactly what my testimony would be. We discussed a 11 number of issues that bear on that having to do with 12 limits and so on, and I think I have given you an 13 idea of the nature of the analysis I have done and 14 how that would affect my testimony. 15 Q. Well, from what I can recollect of what 16 you testified to here today -- or today and yesterday 17 -- was the calculations, you didn't believe the 18 calculation of the limits and levels were defensible? 19 A. Yes, and that certainly is one aspect. 20 Q. I don't remember any other aspects, and if 21 you can expound on -- obviously there was another 22 one, there was data. 23 A. That is related to the defensibility of 24 the limits. 25 Q. That's why I didn't think that was part of 393 1 it. Is there another area in this? 2 A. The sample size is certainly one major 3 issue, the comparability of data in different time 4 periods, sampling methods both in the field and in 5 the laboratory is an issue; lack of relationship 6 between the loads to the WCAs, and in particular 7 Loxahatchee and the application of the limits, is an 8 issue. So there are three. 9 Q. Don't all those things just go to the 10 defensibility of the limits and levels that you were 11 mentioning, sample size, comparability of the data, 12 lack of relationship between loads in the WCAs and 13 the Loxahatchee? It is like a subheading, isn't it? 14 A. Subheading of what? 15 Q. Of the defensibility of the limits and 16 levels? 17 A. Those are all relating to the limits and 18 levels, yes. 19 Q. What I'm trying to get to is, is there 20 anything else you are challenging or if you will 21 testify to other than the defensibility of the limits 22 and levels within the WCAs and the Everglades 23 National Park? 24 A. That I don't know. 25 Q. As of now, what do you believe you will 394 1 testify to, anything else? 2 A. As of now, I do not see anything else. 3 However, given that I have expertise in a number of 4 areas that are listed there, if any of those areas of 5 expertise came to be related to questions having to 6 do with the appropriateness of the -- I'm sorry, you 7 have to read me the exact clause you read me that we 8 started talking about, it had to do with water 9 quality standards. 10 Q. Why don't you read No. 20. 11 A. Whether the phosphorus limiting standards 12 for the EPA or necessary and will prevent violations 13 of state water quality standards. 14 I have been identified as having expertise 15 in a number of areas. You have asked whether there's 16 anything else related to this that I might testify 17 to, and the only statement I can make is that it is 18 not clear to me what other areas there might be now. 19 However, if issues come up pertaining to 20 those areas of expertise that I have been listed as 21 associated with, that have to do with item 20, that 22 are different from those that I gave you, I might 23 offer testimony on them. I cannot foresee what those 24 other areas might be at this time. 25 Q. You would make a good attorney with 395 1 disclaimers and everything. 2 MR. FITZGERALD: I would be mortally 3 insulted. 4 MR. BLANK: He did take care to identify 5 himself as not being a statistician. 6 BY MR. LOREDO: 7 Q. Moving onto the next issue that you are 8 listed under, page 26. Whether the SWIM Plan interim 9 and long-term phosphorus concentration limits are 10 arbitrary, practical, achievable or ecologically 11 significant. 12 A. For me, you can strike the "ecologically 13 significant." I am not an ecologist, so I can't 14 testify to that. Arbitrary, practicable and 15 achievable, yes, there's testimony I would offer on 16 those. 17 And in fact, it is related to the issues 18 we talked about before as to the defensibility of the 19 limits, the manner in which they were specified, the 20 data that was used, and so on. 21 Q. How about the achievability and the 22 practicality of the SWIM Plan? 23 A. Achievability is something I might well 24 testify to. 25 Q. What would you say? 396 1 A. The testimony might well have to do with 2 the likelihood of achieving the limits given 3 assumptions about variability in the data and STA 4 performance. 5 Q. Would the limits be achievable? 6 A. That's not a question that can have a yes 7 or no answer. There are some aspects of the limits 8 that have not been defined and have to do with how 9 often violations might occur and might be allowed to 10 occur; whether they always have to be met; how the 11 sampling would be done, and all of those are issues 12 that I would be prepared to offer testimony on. 13 Q. I got the last part. Can you repeat what 14 you started saying -- actually you can tell me again. 15 (Record read) 16 BY MR. LOREDO: 17 Q. Are the SWIM Plan limits practical? 18 A. I don't think that's something I can give 19 a yes or no answer to. 20 Q. The same variables? 21 A. Practical with respect to what, on what 22 basis? 23 Q. Best way to do it, most practical way to 24 do it, based upon all available data? 25 A. The best way to do what? 397 1 Q. Determination of limits. 2 A. Consistent with the legislative intent? 3 Q. Yes. 4 A. I'm not prepared to testify as to 5 legislative intent. 6 Q. We know what the purpose is, we are trying 7 to get a cleaner Everglades. Would the limits that 8 are being imposed do that? 9 A. One would have to define what one meant by 10 "cleaner." What I would be prepared to testify to 11 would be given the interpretation, the technical 12 interpretation of legislative intent, which is beyond 13 my area of expertise. 14 Would the specific method of applying the 15 limits result in their being met and in what context 16 probabilistically might they be met or not met given 17 assumptions, for instance, STA performance, I would 18 be prepared to testify on that. 19 Q. And if the question was posed whether the 20 limits as imposed would have a positive effect on the 21 Everglades, what would your response be? 22 A. It is not my area of expertise. 23 Q. Then your testimony is going to be limited 24 to the limits that are imposed or the calculation of 25 limits are not defensible? 398 1 A. Not necessarily. 2 Q. Can you expound then? 3 A. The comment was whether I would testify as 4 to whether -- I am paraphrasing you, so maybe you 5 will repeat it -- as to whether the limits would 6 result in a cleaner Everglades. 7 That in my view is an 8 ecological-biological question that I would not be 9 prepared to offer testimony on. If one asked that 10 question in terms of concentrations or other physical 11 measures, then I would be prepared to offer some 12 interpretation and opinion of whether that would 13 happen. 14 Q. Well, you understand that limits have been 15 imposed. Would that affect the concentration within 16 the Loxahatchee? 17 A. That is something I would be prepared to 18 offer some testimony on related to physical analysis 19 of the relationship or lack thereof between the 20 concentrations going into the Loxahatchee and the 21 locations at which the data for application of the 22 limits are or would be collected. 23 Q. What would your testimony be, would there 24 be a relationship from the inflow? 25 A. I think we have indicated that that work 399 1 is in progress. 2 Q. But I think you have stated a preliminary 3 opinion that you thought there was no relationship? 4 A. The preliminary opinion is that we don't 5 see a relationship. 6 Q. Do you have any final -- I think we talked 7 about this -- do you have any opinions now as to what 8 the level should be in the Loxahatchee? 9 MR. BLANK: Levels of what, counsel? 10 MR. LOREDO: Phosphorus levels. 11 MR. BLANK: Standard limits? I am 12 confused about the word "levels." 13 MR. LOREDO: Limits and levels, I kind of 14 use them synonymously -- except when I say "levels," 15 I mean more in the marsh levels. 16 THE WITNESS: Are we talking about levels 17 as in stage or are we talking about phosphorus 18 concentrations? You mean phosphorus concentrations? 19 BY MR. LOREDO: 20 Q. I do mean phosphorus concentrations. 21 A. The question was? 22 Q. What your testimony would be as to the 23 effect within the Loxahatchee in terms of the 24 phosphorus concentration based upon the limits that 25 are being imposed within the SWIM Plan? 400 1 MR. BLANK: I will object to the form of 2 the question. 3 You can answer it if you can. 4 A. The limits as defined in the SWIM Plan 5 apply to the interior marsh stations, so the proposed 6 final limits do -- in fact, the interim apply in part 7 as well, so I don't understand the question. 8 Q. I think I touched upon this yesterday, but 9 let me just double back. 10 Are you going to offer any testimony or 11 opinions as to the 50 parts per billion limit that's 12 being imposed? 13 A. As to whether that limit is? 14 Q. Affects -- well, because one of the 15 rationals under the SWIM Plan is we want to limit the 16 phosphorus coming in because we believe it is 17 affecting the phosphorus within the WCA/1's all the 18 way down to the Park? 19 A. Right. And I have indicated that there is 20 analysis in progress, and the preliminary opinion is 21 that there's not a relationship. I would be prepared 22 to testify to that. 23 Q. That would be your testimony? If you were 24 testifying today, you would testify there's no 25 relationship? 401 1 A. I would certainly testify to that, yes. 2 Q. Do you think that testimony could change 3 with the work in progress? 4 A. Of course, it could. 5 Q. I will read 34 for the benefit of the 6 court reporter. 7 One of the other issues that you have 8 been designated to respond to or to testify to is 9 whether the SWIM Plan arbitrarily establishes 10 phosphorus concentration limits for the Park and WCAs 11 which are below the limits necessary to prevent state 12 water quality violations. 13 Will you offer testimony concerning that 14 issue? 15 A. The testimony with respect to that issue 16 again would have to do with limits and their 17 appropriateness, the appropriateness of inferring 18 both baseline conditions and the relationship of data 19 that have been collected more recently to those 20 baseline conditions. 21 Q. Your testimony would be limited to that? 22 A. There are perhaps other aspects which are 23 related having to do with trends in the Park. 24 Q. I don't think I saw in any of the reports 25 we reviewed today, any trends in the Park. 402 1 A. No, I think I indicate there was some work 2 in progress; you have some computer files. 3 Q. Okay. It is in the computer files? 4 A. I believe there's some work there, and 5 there's also some work in progress that we discussed 6 yesterday. 7 Q. Is this work you are overseeing for 8 somebody? 9 A. No. 10 Q. This is your individual work? 11 A. Yes. 12 Q. Just to summarize, what trend did you find 13 in the Park? 14 A. Well, I reanalyzed William Walker's data, 15 and I think we discussed this yesterday, and -- 16 Q. We might have. 17 A. -- there seemed to be some indication of 18 uptrend in the period he had analyzed through 1989, 19 and I think we had just discussed that based on the 20 analysis of Dr. McClave, there's some indication of a 21 change in direction of that trend since then, and 22 that we have been attempting for some some time, and 23 apparently just in the last week and a half or so 24 have obtained more recent data in the last four years 25 and have to incorporate it into the analysis. 403 1 Once I have completed that analysis, I 2 would then be in a position to issue an opinion on 3 that. 4 Q. Are you working concurrently with Dr. 5 McClave or is it exclusive? 6 A. I'm sorry? 7 Q. Well, you had indicated Dr. McClave is 8 working on a similar analysis in the Park? 9 A. Yes. 10 Q. When you met with him and have shared some 11 information -- are you saying you are both working on 12 similar analysis or different analysis? 13 A. There certainly is some overlap. 14 Q. Where would your analysis differ or where 15 is your analysis going to differ? 16 A. I don't recall exactly what he had done. 17 His preliminary analysis seemed to be using 18 parametric methods. My preference is to use 19 nonparametric methods. That's one aspect. 20 I don't recall which independent variats 21 he might have adjusted for. My inclination is to 22 adjust or account for flow variation. 23 Q. Anything else? 24 A. Those are two that come most immediately 25 to mind. 404 1 Q. Does anything else come to mind? 2 A. No. 3 Q. The last issue you have been designated as 4 -- obviously this could be subject to change -- is 5 whether the SWIM Plan's interim and long-term 6 phosphorus concentration limits for the Park and WCAs 7 have been correctly determined by the District. 8 A. That's fairly closely related to the issue 9 we just discussed. 10 Q. Would your testimony be they are closely 11 related? 12 A. I'm sorry? 13 Q. I'm sorry, that they were correctly 14 determined by the District? 15 A. Most likely it would be that they were 16 incorrectly determined. 17 MR. LOREDO: I will rest and turn you over 18 to my counsel here, co-counsel here, Tom Fitzgerald. 19 CROSS-EXAMINATION 20 BY MR. FITZGERALD: 21 Q. It is virtually impossible, and we will 22 not finish today, I'm sorry. It is already pushing 23 4:00, and we will keep going, and I am confident we 24 will be done before noon tomorrow. 25 A. Whatever it takes. 405 1 Q. Doctor, for the record, I'm Assistant U.S. 2 Attorney Tom Fitzgerald representing the United 3 States in this pending series of administrative 4 actions for the Division of Administrative Hearings 5 here in Florida. 6 I would like to go back and ask you just a 7 few follow-up questions on your designation of 8 witnesses that appeared, or areas of witness 9 testimony as it appeared in the February 4th, 1994 10 pretrial disclosure of issues and witnesses filed by 11 the Florida Sugar Cane League and U.S. Sugar 12 Corporation. 13 Counsel asked you some questions about 14 them, so I won't go into any great detail, but 15 several of those discussed the appropriateness in one 16 character or another of the interim, long-term 17 phosphorus concentration limits and levels for 18 Loxahatchee National Wildlife Refuge and Everglades 19 National Park that are contained in the SWIM Plan 20 from the perspective of their appropriateness or 21 efficacy under Florida water quality law. 22 What is your understanding of Florida 23 water quality law as it would factor into your 24 testimony in this matter? 25 A. I wouldn't be testifying on Florida water 406 1 quality law. 2 Q. So I take it, then, it is fair to say that 3 you will not opine as to the appropriateness of those 4 limits and levels, either interim or long-term, as 5 they may in fact define or not define the levels 6 required to meet the nondegredation standard of 7 Florida water quality law? 8 A. I think I indicated that insofar as 9 Florida water quality law and nondegredation 10 standards and so on are reduced to technical 11 criteria, I would be prepared to offer testimony as 12 to their appropriateness. 13 Q. And as I understand it, your testimony to 14 their appropriateness would be based essentially on 15 your analysis of their achievability or the 16 appropriateness of the assumptions upon which they 17 were based? 18 A. And/or whether they in fact represent what 19 they claimed to represent, for instance, certain 20 aspects are based on certain probability levels. Are 21 those in fact properly represented technically? 22 Q. So you have no opinion on what the correct 23 standard for phosphorus in the Everglades ecosystem, 24 the EPA, should be under Florida law? 25 A. As related to ecological factors, no. 407 1 Q. As related to statistical analysis based 2 on all available data you have reviewed? 3 A. No, that is not a correct statement you 4 have just made. 5 Q. You have such a number in mind then? 6 A. I did not say I had a number in mind, I 7 said I was maybe prepared to testify as to 8 appropriateness on a statistical grounds, but on 9 ecological grounds, I would not. 10 Q. On statistical grounds, what number do you 11 believe is appropriate to meet the requirements of 12 Florida law for the Everglades Protection Area? 13 A. There is no particular number that I would 14 be in a position to state at this point. 15 Q. Do you plan to conduct additional analysis 16 to identify a number of alternatives to those 17 contained in the SWIM Plan at issue in the case? 18 A. I do not at present intend to do that. 19 Q. Have you discussed with counsel or any of 20 the other consultants in this matter deriving such an 21 alternative number? 22 A. No. 23 Q. Have you done any work thus far 24 specifically intended to derive such a number? 25 A. I think you are couching the question in 408 1 terms that are not necessarily consistent with the 2 type of analyses I have been doing, and to my 3 understanding of your question, the answer is no. 4 Q. One of the designations specified is that 5 you would testify as to whether the SWIM Plan 6 arbitrarily establishes phosphorus concentration 7 limits for the Park and WCAs -- 8 A. I think what you need to understand is 9 that those are designations that are made in legal 10 terms, that until they are reduced to technical terms -- 11 Q. I'm sorry to interrupt you, but there's no 12 question now. Let me finish the question and then 13 explain to me. You may find that the explanation of 14 the line you are pursuing has nothing to do with what 15 I'm asking. 16 The balance of that -- which are below the 17 limits necessary to prevent state water quality 18 violations. So as that is drafted, I understand the 19 arbitrary concern, and I think counsel and you 20 addressed that. 21 I am not asking you about whether the SWIM 22 Plan was arbitrary from a statistical analysis sense, 23 I think you fairly well established it probably is, 24 or that would be your current testimony or 25 understanding, but am I correct that you are not 409 1 intending to address in any way whether those limits 2 are lower than the limits necessary to prevent state 3 water quality violations? 4 A. I cannot answer that question at present 5 without you defining for me state water quality 6 standards and how they related to the criteria. 7 Q. So you don't know what the criteria are? 8 A. I know what the proposed limits are, I 9 don't know the numerical state water quality 10 standards, no. 11 Q. Counsel in his letter of March 1st 12 identified as one of the sources of your knowledge as 13 something you reviewed and may rely, the Everglades 14 SWIM Plan of March 13th, 1992 which are comprised of 15 three volumes. 16 Have you read the entire SWIM Plan? 17 A. I have not read it cover to cover, no. 18 Q. Do you understand the nature of the 19 document as a planning document intended to be 20 modified over time? 21 A. I'm not sure I understand your question. 22 Do I understand every line of that document? 23 Q. What is your understanding of the nature 24 of the document and its intended use over the course 25 of time in Florida by the Water Management District? 410 1 A. I haven't made a judgment on that. My 2 expertise has been offered with respect to particular 3 aspects of that plan, and I think we have talked 4 about particular appendices. 5 Q. Other than appendices E and F, what 6 portions of the documents comprising the plan have 7 you reviewed in detail? 8 A. E and F were the major parts of it. 9 Q. Did you review any of the discussion in 10 the planning document wherein the difficulty with 11 identifying currently a specific numeric standard 12 under Florida water quality law for phosphorus in the 13 Everglades Protection Area was discussed? 14 A. Have I read or have I had discussions? 15 Q. Have you read the portions of the plan? 16 A. I don't recall. 17 Q. The next question is, in your 18 conversations with Dr. Davis, counsel and others, has 19 that issue been discussed with you? 20 A. There have been aspects of that discussed, 21 yes. 22 Q. What is your understanding of the current 23 status of the Florida law with regard to numerical 24 standards for phosphorus entering the Everglades 25 Protection Area? 411 1 A. That is not something that is an area that 2 I have been asked to comment on. 3 Q. The question is, what is your 4 understanding of the current status of Florida law 5 with regard to a numeric standard for phosphorus 6 entering the EPA? 7 A. That I don't recall at this point. 8 Q. So you have no current understanding then? 9 A. That is correct. 10 Q. If you have no current understanding of 11 what that requirement is under Florida law, how do 12 you plan to offer testimony on the issue of whether 13 the SWIM Plan's phosphorus limiting standards will 14 prevent violations of state water quality standards? 15 A. I think -- 16 MR. BLANK: I will object to the form of 17 the question, counselor. I don't know whether you 18 are inquiring concerning the standard or concerning 19 the OFW criteria? Your previous questions were 20 termed in the nutrient standard, and now you seem to 21 be asking something else. 22 MR. FITZGERALD: That's right. I am 23 referring to the designation of witness paragraph 20 24 that counsel previously read which says whether the 25 SWIM Plan's phosphorus limiting standards for the EPA 412 1 are necessary to or will prevent violations of state 2 water quality standards. 3 BY MR. FITZGERALD: 4 Q. The state water quality standard for EPA, 5 you have just testified you have no current 6 understanding of what that requirement is? 7 A. That's not what I said. 8 Q. Okay. Let's go back, what is your 9 understanding of the state water quality standard 10 requirement? 11 A. My understanding of state water quality 12 standard is a nondesignation standard as pertains to 13 the OF W. 14 Q. What about class 3 standards, do you 15 understand their implications? 16 A. I do not have a firm understanding of all 17 of the details of Florida law. 18 Q. What is your understanding of whether the 19 limiting standards for the EPA contained in the SWIM 20 Plan will or will not meet that nondegredation 21 standard? 22 A. My understanding is that there's a certain 23 amount of interpretation as to how that's to be 24 applied and what nondegredation is to mean, and a 25 certain amount of interpretation is present in the 413 1 SWIM Plan proposals. 2 Q. Based on your understanding of the SWIM 3 Plan's limiting standards for the EPA, do you believe 4 you can offer an expert opinion on whether those 5 limiting standards will prevent violations of state 6 water quality standards? 7 MR. BLANK: Objection to the form of the 8 question. 9 THE WITNESS: Are you waiting for me? I 10 think he just had an objection. 11 MR. BLANK: You can go ahead and answer. 12 MR. FITZGERALD: You are required to 13 answer unless your counsel directs you not to. 14 That's preserved for the record later. 15 THE WITNESS: Well, I think I made it 16 fairly clear as to what the nature of my expertise 17 and testimony would be. The nature of my expertise 18 and testimony is not to how the Florida law is 19 encoded into certain standards, it is to whether or 20 not the standards as they are encoded in a technical 21 statement are appropriate statistically and whether 22 or not it is likely that those technical statements 23 can be met. So that is the aspects that I can 24 comment on. 25 Whether or not the technical statement 414 1 would be consistent with the narrative in the law, is 2 a question which is outside my realm of expertise. 3 Q. You testified yesterday that Lisa Dally -- 4 A. Dally Wilson. 5 Q. Dally Wilson did the actual, if you will, 6 number crunching, the running of the kriging programs 7 for you? 8 A. That's correct. 9 Q. She was a subcontractor? 10 A. That's correct. 11 Q. Who chose kriging as the analytical method 12 for reviewing the data that was provided to you? 13 A. Well, that had already been done in the 14 Reddy reports. 15 Q. Yes, but you took the kriging method and 16 applied it in a number of areas that Reddy did not? 17 A. Yes, that was my decision that I concurred 18 with the earlier work, that it was in fact an 19 appropriate method, although I might have disagreed 20 with some of the decisions that were made in its 21 application in the earlier work. 22 Q. Prior to being contacted in the summer or 23 fall of '92, if I have the date right, had you had 24 any other prior experience in the Everglades 25 ecosystem? 415 1 A. In the Everglades, no. Well -- yes, the 2 Everglades as it is commonly defined, the answer is 3 no. 4 Q. Why do you draw that distinction, what 5 area? 6 A. Well, as you aware from our discussions 7 yesterday, I have done work for the District 8 associated with Lake Okeechobee, and in fact, there 9 are some aspects of that work that had to do with 10 loading estimates and with sampling at inflow and 11 discharge points from the lake. 12 Some of those happened to be the same 13 structures that we are discussing now which are in 14 flow points to the EAA. However, those were of minor 15 concern in that work so it is peripheral. 16 Q. The S-2, S-3, those structures? 17 A. Exactly. There are samples collected as 18 part of Lake Okeechobee program at those structures 19 and so those were included, although the emphasis in 20 the earlier work much more had to do with the north 21 side of the lake. 22 Q. What was your conclusion on the 23 relationship between flow and concentration in the 24 Lake Okeechobee work? 25 A. In Lake Okeechobee, it is quite consistent 416 1 with what we discussed yesterday for structures 2 further south, that the relationships were weak, that 3 it was sometimes a weak-positive relationship or 4 increased in concentration and flow, and in many 5 cases there was none. 6 Q. Since you completed that work, before you 7 technically started this work, did you bring that 8 conclusion with you, did it affect your view or your 9 analysis of the relationship in structures strictly 10 discharging south from the EAA into the EPA? 11 A. Well, let's say I wasn't surprised, but it 12 wasn't, the analysis wasn't conducted sort of in 13 search of data to justify a conclusion. It really 14 makes no difference to the type of analysis that I'm 15 doing whether there is a relationship or is not. 16 Q. Does your subcontractor, to your 17 knowledge, have any experience in the Everglades 18 ecosystem? 19 A. Prior to this work? 20 Q. Yes. 21 A. I don't believe so. 22 Q. And other than the Lake Okeechobee work, 23 then -- 24 A. I'm sorry, to clarify that. To the extent 25 to which I had previous experience peripherally in 417 1 Lake Okeechobee where she had worked on that project 2 as well, the same statements would pertain to her 3 experience. 4 Q. Other than the two small stormwater 5 treatment areas you described yesterday with respect 6 to her master's program thesis, do you know if Ms. 7 Wilson has any experience with major wetland 8 treatment systems? 9 A. I'm not aware of that, but I couldn't 10 speak absolutely yes or no. Probably not. 11 Q. How about you, the same question? 12 A. Definitely not. 13 Q. I qualified it by saying "major." Let's 14 be a little more expansive. Do you have any 15 experience in analysis, design, construction, 16 monitoring plans, across the board, anywhere within 17 your field of expertise with stormwater treatment 18 systems, man-made marshes? 19 A. Man-made marsh systems as opposed to 20 stormwater treatment systems because you are quite 21 aware that I have experience with stormwater 22 treatment systems. The answer to marsh systems, no. 23 Q. Do you have any experience with 24 constructive wetlands of any sort whatsoever? 25 A. No. 418 1 Q. Now, with regard to stormwater treatment 2 systems, those with which you are familiar, are 3 essentially municipal waste treatment systems? 4 A. No, they are not waste treatment systems. 5 Q. Municipal stormwater runoff treatments? 6 A. Stormwater runoff treatment systems, yes. 7 Q. Does the kriging analysis that you caused 8 to have performed and on which your reports are 9 based, have any application to those types of 10 systems? 11 A. No. 12 Q. Have you ever applied such kriging 13 analysis to any natural system? 14 A. Yes, I think you are aware of the 15 publication record. 16 Q. And those systems were riverine systems? 17 A. No. 18 Q. I must have missed it. 19 How would you characterize the systems? 20 A. The demonstration application in the paper 21 in question, I think I indicated yesterday, was a 22 groundwater quality application. 23 Q. That was in the demonstration project 24 beyond for your Ph.D? 25 A. No. That was a paper that was published 419 1 in Water Research in 1981 and co-authored by James 2 Hughes. It was based in large part on work he 3 conducted as part of his master's thesis. 4 Q. That was a student who was under your 5 guidance? 6 A. Yes. 7 Q. Other than that circumstance, in what 8 other types of systems have you applied that sort of 9 analysis? 10 A. Applied kriging? 11 Q. Yes. 12 A. Have I personally or have I been involved 13 in projects that have investigated it? 14 Q. Personally. 15 A. Personally, push the numbers through, 16 ground out the answers? None. 17 Q. You said yesterday that your appointment 18 is primarily a teaching and research appointment. 19 What does that mean? 20 A. That means virtually all of my support 21 from the University of Washington is derived from 22 grants and contracts, support that I bring in. 23 Q. Does any of your professio