Project Update: 9/29/11
Activities to address the goals of the SWG funded project were initiated in March 2011.
To date, the following activities have been the primary focus: 1) literature review, 2) analysis of gage information and period of record for each gage, 3) training and use of flow statistics software, 5) reviews of Bayesian mixed modelling and random forest techniques, 5) outreach and additional collaboration, and 6) summer 2011 sampling. These focal areas are briefly summarized here.
Numerous key studies regarding stream classification, streamflow modelling, and environmental flows were recently published and a few are reviewed here. Richter et al. (2011) reported a presumptive standard for environmental flow protection using the sustainable boundary approach (Richter 2009), which sets limits of augmentation and depletion (essentially tracking-curves) along normal baseline hydrographs. Using random forest modelling techniques (i.e., Cutler et al. 2007), Carlisle et al. (2010) modeled flow across the conterminous U.S. using 39 landscape and climate predictor variables. McManamay et al. (2011) classified streams in Mid-Atlantic States using HIT software (Henricksen et al. 2006) and K-means clustering, and found that classification generally were concordant with ecoregional classifications. Liermann et al. (2011) investigated eflows for Washington State using both Bayesian mixture modelling and random forest modelling, in addition to assessments of geomorphic influences on hydrologic predictions and assessments of different climate change scenarios on hydrologic predictions. These methods of classification and modelling of Liermann et al. (2011) will be followed more or less in our Arkansas project. A different modelling approach was that of Larned et al. (2010), where ELFMOD was used to assess longitudinal flow variation, which uses spot gaging data (e.g., indication of wet or dry or wetted margin width) and time series data to predict flow magnitude along river segments. This modelling approach might be useful for assessing longitudinal flow variation in Ozark streams, especially in cases where flow variation (and permanence) is influenced by springs and groundwater interaction (e.g., Dry Fork Creek in Madison County and Bear Creek in Searcy County).
Periods of records (POR) for gages in Arkansas were finalized, resulting in 177 stream gages with > 10 year POR for a time period of 1950 – 2009. For these gages, daily mean files and peak flow files were compiled from NWIS, for a total of 354 files. The Hydroecological Integrity Assessment Process package (Henricksen et al. 2006) was initially chosen and used to calculate flow statistics for each gage. However, because of some problems running this software, alternative methods for calculating flow indices were sought. Konrad (2011) published the Environmental Flow Allocation and Statistics Calculator, an Excel macro that calculates flow indices and additionally provides different scenario settings for directly manipulating environmental flow allocations for particular gages. Flow allocation can be related to different bioperiods for assessment of specific critical flow scenarios. The Eflow calculator has been easy to use and flow metrics have been successfully calculated, also requiring pre-processing of only one file (i.e., daily means). Using this calculator, hydrologic indexes for gages in Arkansas have been calculated and a 158 site x 43 flow variable matrix developed for stream classification analysis. Furthermore, the Eflow calculator has previously been used in a broad- scale study to determine hydrologic classifications of rivers and streams throughout the southeastern U.S. (C. Konrad pers. comm., in manuscript).
For determining hydrologic baseline (reference) conditions, we are using the data of Falcone et al. (2010) to select reference gages based on landscape variables. While this environmental data is extremely thorough and useful for reference condition determinations and predictive modelling, some AR gages that had < 20 yr. POR did not have this environmental attribute information. Therefore, J. Falcone was contacted and he very kindly offered to calculate the same environmental attributes for our additional gages not occurring in GAGES and with less than 20 yr. POR (n = 58). Furthermore, most of this environmental data in Falcone et al. (2010) will be used as predictors in random forest modelling of hydrologic conditions.
We have made contact with several researchers throughout the U.S. that are involved with environmental flow projects. Discussions have occurred with M. Davis, J. Faustini, C. Konrad, J. Falcone, M. Kennard, and R. McManamay. A phone conference was held on September 16th with Dan M. and Scott L, Del Lobb, Paul Blanchard, Jason Persinger and Emily Tracy-Smith of Missouri Department of Conservation, Shannon Brewer (Oklahoma Cooperative Fish and Wildlife Research Unit), and Jeff Quinn (AGFC). The goal of the phone conference was to determine the status of different eflow projects in Arkansas and Missouri. Missouri has completed stream classifications (Kennen et al. 2009), with current activities (since March 2011) including compilation of literature dealing with ecological responses to pertinent flow characteristics. Based on the phone discussions, Dan and I decided to include the Missouri gages located within the Ozark Plateau, which will help with stream classifications and likely expand the universe of potential sites for assessing ecological responses across flow altered and reference sites within stream classes. Missouri folks welcomed the collaboration and will be providing us site information for gages used in stream classifications for the Ozarks. The past two weeks have consisted of downloading gage information and pre-processing files for gages in Missouri and in four ecoregions of Oklahoma (Ouachita, Boston Mountains, Arkansas Valley, and the Ozark Highlands). Adding these Oklahoma gages does not significantly increase the work load yet will provide additional sites to help reduce model uncertainty in stream classifications, with the obvious constraint of a priori ecoregional classification (classification using gages across these four adjacent OK ecoregions should reduce this ecoregional constraint because all gages will be classified simultaneously, i.e. hydrologic variation might overlap between streams in adjacent ecoregions). Current tasks involve pre-processing Missouri and Oklahoma NWIS files (i.e. filling empty cells and removing provisional/unapproved data) as required by the Eflow statistics calculator.
In summer 2011, our new Ph.D. student, Dustin Lynch, and crew (N. Vogt and J. Schluterman- Arkansas Tech, UA undergraduate Toshiki Hayashi, and Brie Olsen-Coop Technician) sampled fish and macroinvertebrate communities in the Little Red River watershed at sites previously established by TNC. Biological information from this work will be related to existing hydrologic and habitat data to comprise a pilot eflow project associated with ungaged sites. Sampling in the Saline watershed was attempted but because of drying at numerous sites communities there were sampled at only a few locations.
In October – November 2011, we will be conducting Bayesian mixture modelling and potentially other stream classification techniques. Concurrent with these activities include preparing existing environmental attribute data and acquiring/calculating new environmental data for determining reference gages and for random forest modelling to predict hydrologic variation. In late fall/winter 2011, we will begin processing existing biological data and developing methods to analyze flow-ecology responses for selected streams of the Ozark Plateau. During this time we will begin determining effective numbers of reference and flow-altered reaches/streams in the Ozarks for establishing the field sampling design for fish and macroinvertebrate communities.
We are open to all comments regarding this summary of the current activities for the eflow project. Please don’t hesitate to call or write anytime with questions or issues of concern.
Literature cited
Carlisle, D. M., J. Falcone, D. M. Wolock, M. R. Meador, and R. D. Norris. 2010. Predicting the natural flow regime: models for assessing hydrologic alteration in streams. Rive r Research and Applications, 26:118 – 136.
Cutler, D. R., T. C. Edwards, K. H. Beard, A. Cutler, K. T. Hess, J. Gibson, and J. J. Lawler. 2007. Random forests for classification in ecology. Ecology, 88:2783 – 2792.
Falcone, J. A., D. M. Carlisle, D. M. Wolock, and M. R. Meador. 2011. GAGES: A stream gage database for evaluating natural and altered flow conditions in the conterminous United States. Ecology, 91:621.
Henricksen, J. A., J. Heasley, J. G. Kennen, and S. Niewsand. 2006. User’s manual for the hydroecological integrity assessment process software (including the New Jersey Assessment Tools): U.S. Geological Survey, Biological Resources Discipline, Open File Report 2006-1093, 71 p.
Kennen, J. J. A. Henricksen, J. Heasley, B. S. Cade, and J.W. Terrell. 2009. Application of the Hydroecological Integrity Assessment Process for Missouri Streams, U.S. Geological Survey Open-File Report 2009-1138, 57 p.
Konrad, C. P. 2011. Environmental flow allocation and statistics calculator: U.S. Geological Survey Open File Report 2011-1166, 46 p.
Larned, S. T., D. B. Arscott, J. Schmidt, and J. C. Diettrich. 2010. A framework for analyzing longitudinal and temporal variation in river flow and developing flow-ecology relationships. Journal of the American Water Resources Association, 46:541 – 553.
Liermann, C. A. R., J. D. Olden, T. J. Beechie, M. J. Kennard, P. B. Skidmore, C. P. Konrad, and H. Imaki. 2011. Hydrogeomorphic classification of Washington state rivers to support emerging environmental flow management strategies. River Research and Applications, DOI: 10.1002/rra.1541.
McManamay, R. A., D. J. Orth, C. A. Dollof, and E. A. Frimpong. 2011. A regional classification of unregulated stream flows: spatial resolution and hierarchical frameworks. River Research and Applications, DOI: 10.1002/rra.1493.
Richter, B. D. 2009. Re-thinking environmental flows: from allocations and reserves to sustainable boundaries. River Research and Applications, 25:1 – 12.
Richter, B. D., M. M. Davis, C. Apse, and C. Konrad. 2011. A presumptive standard for environmental flow protection. 2011. River Research and Applications, DOI: 10.1002/rra.1511.