Following are excerpts from the comments of Dr. Sam Upchurch on the draft of the proposed Minimun Flows and Levels (MFLs) written by the Suwannee River Water Management District (SRWMD.) Dr. Upchurch is an internationally respected scientist in Florida regarding hydrogeolgy and karst.
The latest and current MFLs were set in 2015 and they are updated regularly. The setting of the MFLs is a guide for allowing pumping from the river and the protocol is to take the most possible without causing “significant harm” to the river. This, of course, is operating backwards, as the mandate of our agencies is to protect, not squeeze out as much as possible without killing the river.
But that is another long story.
The MFLs are set by using a model to predict what will happen to the flow of a river given certain factors. To be useful, information introduced into the model must be accurate and adequate. The model must also be calibrated to the characteristics of the region in which it is employed.
The water scientists at SRWMD request input from a Peer Review Panel which in this case consists of three other scientists, and they also welcome comments from the public. The Alachua County Environmental Department engaged the services of Dr. Upchurch to read and comment on the draft. Here we might add that the Peer Reviw Panel, consisting of Drs. William Dunn, Louis Motz and Adam Munson, were quite negative in tone. Of the 177 items in the draft, the Panel took issue with over half. They can be seen at this link.
Dr. Upchurch continues the harsh but necessary criticism of this draft which, if finalized, may result in reclassifying the Santa Fe as restored and in health and thus, ready to give up more water.
We give thanks to Merrillee Malwitz-Jipson for distributing Dr. Upchurch’s comments.
Comments by OSFR historian Jim Tatum.
– A river is like a life: once taken,
it cannot be brought back © Jim Tatum
SAM UPCHURCH P.G., PH.D.
Dr. Sam Upchurch’s report was submitted as an attachment to the cover letter from Alachua Coiunty Environmental Protection Department (ACEPD) (ACEPD 2020).
Dr. Upchurch’s report (Upchurch 2020) is a formal technical peer review of the proposed MFLs. He places emphasis on (1) the ability of the proposed MFLs to protect these water bodies from “significant harm” as directed in Chapter (§) 373 Florida Statutes and (2) the utility of the MFLs for water-resource management.
SU 1—page 1, first bullet
The MFL basis report, dated December 2019 and prepared by HSW Engineering, Inc., deals with the appropriate issues for MFL development. The report develops the basis for revising MFLs set in 2013. The 2013 MFL basis document concluded that water had been over-committed and concluded that significant harm had occurred to the rivers. As a result, a Recovery Strategy was required. Part of the Recovery Strategy included development of a new, regional groundwater flow model to replace the one used in 2013.
SU 2—page 1, second bullet
Chapter 373 F.S. requires that “best available data” be used in development of the MFLs. The basis report includes the best available data for the Lower Santa Fe River (LSFR) and Ichetucknee River (IR,) but not for the Priority Springs (PS). Use and manipulation of these data are a subject of this peer review, however.
SU 3—page 1, fifth bullet The report incorporates several new approaches to MFL development. These include use of the new groundwater flow model (the North Florida and Southeast Georgia [NFSEG] model); use of the System for Environmental Flow Analysis (SEFA), a method for relating riverine flows to such issues as fish passage; and reliance on a Reference Time Frame (RTF) hydrograph, which was used to propose theoretical hydrographs for the rivers assuming that no groundwater withdrawals were present.
SU 4—page 2, first bullet
As a result of use of these new approaches, the basis document concluded that significant harm has not occurred in the rivers and that water is available for additional use. These conclusions are in contrast to the 2013 MFL even though the 2019 and 2013 MFL critical flows are similar for the IR and the critical flow for the LSFE at the Ft. White gage are lower in the 2019 MFL than in the 2013 MFL.
SU 5—page 2, second bullet
My review of the basis document indicates that the proposed MFLs are not adequately supported by the relevant scientific and technical literature and that large uncertainties in the RTF data, including historical pumping data used in development of the NFSEG model, and thresholds used to evaluate Water Resource Values (WRVs) specified in §373 F.S. with SEFA are not well explained or defended.
SU 6—page 2, third bullet
Several of the socio-politically sensitive WRVs, including recreation and aesthetics are not well developed. These WRVs are often neglected because (1) they are hard to quantify and (2) it is assumed that the other WRVs, which are quantifiable, protect these WRVs. Given the issues that currently are important to stakeholders and often dominate the press; these WRVs should be better evaluated and the opinions of the authors defended.
SU 7—page 2, fourth bullet
Priority springs were thought to be protected from harm by the river MFLs. It is my professional opinion that this is not the case and that the springs are not adequately protected. Furthermore adds Dr. Upchurch: It is recommended that this additional work on spring MFL development methods be initiated immediately. The basis document is weak on references and shows little understanding of specific springs and resurgences or spring behavior. MFLs may be possible for priority springs with a better understanding of spring functions and springshed characteristics. Finally, it would be advantageous to develop a comprehensive review of methods that have been utilized for MFLs of Florida springs in order to assist in development of future MFLs.
SU 8—page 2, fifth bullet
The basis document fails to provide sufficient detail to evaluate conclusions contained in the report. The report is poorly referenced and fails to adequately explain karst processes in the study area.
SU 9—page 2, sixth bullet
The RTF hydrograph was used as a baseline for establishing MFLs. There is no discussion of the reliability of historic pumping data, a critical step in the MFL process. This is a serious shortcoming because errors and uncertainties in estimating the RTF river flows cascade through the entire MFL processes. The following peer review enumerates many of the uncertainties associated with the MFL methods. Uncertainties are cumulative, so by the time a MFL is proposed, there may be consequential possible error.
SU 10-page 2, seventh bullet
The SEFA approach is not well defended and many of the critical thresholds used to determine adequate flow conditions for a WRV are based on personal communications and other “soft” sources. There is no adequate validation of the SEFA method, especially in comparison with the PHABSIM methodology, which was utilized in the 2013 MFL basis document and led to the conclusion that water resources had been over committed.
SU 11—page 2, eighth bullet
Transfer of important, flow-related thresholds from one gage to another by ratio-and- proportion methods needs substantial improvement in terms of justification and precedent.
SU12—page 2, ninth bullet
There is a historic precedent used amongst the water management districts that a 15 percent reduction of flow from baseline conditions represents significant harm. This assumption was used for the 2019 MFLs and needs to be justified with appropriate literature citations.
SU13—Page 2, first bullet at top of page
The MFLs for the rivers consist of one flow value and recurrence interval, namely the median flow that protects all WRVs. This conservative approach assumes that the shape MFL-defined curve. If rigorously followed, there will be no ability to allocate water during high flows or from base flow (flow from springs). It is my opinion that (1) this MFL structure limits water management options and (2) will be difficult to follow. It does, however, make interfacing with a steady state water flow model simple.
SU 14—page 13, first and second bullets
Description of the karst features, including the Cody Scarp, riverine sinks and siphons, and springs is meagre. Few relevant references are cited and there is no significant demonstration of understanding of the hydrogeologic function of the escarpment. The red line shown on Figures 2 and 3 of the main basis report purports to show the Cody Scarp and is attributed to the District. This line does not reflect the scarp, which is a three-dimensional geomorphic feature that begins, on the LSFR, near Worthington Springs and extends westward to just past the River Rise. Springsheds within the Cody Scarp include swallet to spring systems. The spring systems within the Cody Scarp are complex. These complexities are not indicated in the basis document. For example, discharge from the River Rise is largely derived from the river at high discharge and groundwater from the east at low flow.
SU 15—page 13, fifth bullet
The basis report appropriately begins the MFL discussion with the stream gage at Worthington Springs, which is part of the Upper Santa Fe River (USFR) and for which a MFL has been established. The basis document ties the LSFR to this gage but does not cite effects of the USFR MFLs on the LSFR MFLs.
SU 16—page 13, sixth bullet
The USFR MFL basis report anticipated that Olustee Creek, which is downstream from Worthington Springs, would be included in the LSFR MFLs. There have been stream gages on Olustee Creek in the past which are mentioned in the 2019 appendices. Given the proposed phosphate mining in Union and Bradford counties, Olustee Creek flows should be considered in the 2019 basis document. This could range from MFL development based on synthesized data or, at least, discussed in terms of how Olustee Creek relates to the proposed MFL at US 441.
SU 17— page 14
I find the single value MFL based on median RTF flow to be problematic. It would be preferred to see a flow regime specified in the MFL based on multiple WRV thresholds. The single value method that relies upon maintenance of the shape of the flow-duration curve does not allow for permitting exceptions that might make water available under certain circumstances, such as “scalping” of peak flows or permitting of base flow from a spring, both of which would alter the shape of the flow-duration curve and a deviation of the median flow.
SU 18-page 15
Ability to Implement and Practicality of the Proposed MFLs–I have some concerns with this issue. As stated above, there is likely uncertainty with the RTF method. Therefore, identification of available water, or lack thereof, based on the RTF is problematic. Furthermore, use of a single control point at median flow on the RTF flow-duration curve constrains high and low flow conditions in order to maintain the shape of the flow-duration curve. A set of MFLs that specify a regime of flows and recurrence intervals to protect the WRVs affected by high and low flows is preferable, in my opinion, because it allows permitting related to extreme flows as long as the relevant WRV thresholds are not exceeded. Implementation as alluded to in the basis document would suggest that the flow regime is protected by maintaining the shape of the RTF-derived flow-duration curve and simply allowing a shift in median flow and all other points on the flow-duration curve of a specified amount without changing the recurrence intervals at any shifted flow. This will likely be difficult accomplish and uncertainty based on the RTF will be an issue.The shape of the flow-duration curve will be especially hard to maintain if capturing of peak flows for off-line storage and recovery or if base flow is intercepted by permitted water withdrawals from a spring. It is my opinion that reliance on water managers being able to maintain the shape of a flow-duration curve without written guidance is optimistic.
SU 19—references scattered
Problematic uncertainties Were “Best Available” Data Utilized? In terms of hydrological data, yes. See the above discussions of this topic. Literature citations, especially those that characterize the study area and ecological/habitat systems, karst, and uncertainties, are limited in the basis document and appendices Adequacy of Data Used to Support the Proposed MFLs See the comments above. Data are probably adequate and the best available, but they are not fully validated as being adequate or reliable in the basis document or appendices. Statistical methods used to fill data gaps in the stream-flow time series are widely used and considered to meet industry standards. However, discussions of “goodness of fit” and uncertainty were not presented in the basis document. These discussions may be in the appendix. Use of SEFA to establish WRV thresholds is an advancement over PHABSIM. However, the appendix discussing the methods in detail is weak and does not provide confidence to the reader and water manager in the method. The discussion in the basis document should include considerations of applicability to the LSFR and IR, uncertainties, and improvements over traditional methods, such as PHABSIM. WRV discussions are good, but some of the threshold values are based on personal communications or presentations at technical meetings. If possible, peer-reviewed thresholds should be utilized. If the only authoritative threshold source is a personal communication or “gray literature”, any uncertainty should be revealed.
SU 20—page 16
Ability of the MFLs to Define a Hydrologic Regime—Taken verbatim, the MFLs do not specify a regime. A flow regime is maintained only if the MFL is managed so that the shape of the flow-duration curve never changes.
SU 21—page 16
Technical Need for a Recovery Strategy–The 2019 basis document indicates that there is less available water available than the 2013 MFL document allowed in the LSFR at the Ft. White gage and the same water availability in the IR. Based on the allowed flow reductions from the new RTF flow-duration curve, there is water available at both LSFR gages and in the IR (Table 33, 2019 basis document). As a result, there is no indication of the need for a Recovery Plan in the 2019 basis document. The lack of need for a Recovery Strategy should be discussed in the 2019 basis document, especially in consideration of the 2013 MFL basis document which found that water in both the LSFR and JR had been over committed and that a Recovery Plan was merited. The development of a Recovery Plan in 2013 was, in part, a result in uncertainties in the MFL and historic flows after removal of permitted uses.
SU 22—page 16
Approach of the MFLs to Natural Stresses, Such as Climate Change and Natural Hydrological Cycles–
Climate change and hydrologic cycles are mentioned in the basis document, but they are not directly addressed in terms of the MFLs and management of water availability.
SU 23—page 16
Appropriateness of Statistical Methods and Conclusions–The statistical methods are appropriate. There is a need for uncertainty analysis concerning detailed hydro logic data gap filling methods and
development of the RTF data.