Recently Completed CRB R2O Research (post-2016)
Quick Links
CRB R2O Program Homepage
Ongoing CRB R2O Research
Recently Completed CRB R2O Research (post-2016)
2008 – 2016 Research
Publication Abstracts
Colorado River Basin Climate and Hydrology State of the Science Report
Institution(s): Western Water Assessment (WWA), Climate Assessment for the Southwest (CLIMAS), National Center for Atmospheric Research (NCAR), University of Arizona (UA), University of Nevada-Reno (UNR), Lynker Technologies
Years: 2018 – 2020
PI(s): Lukas (WWA), Payton (WWA)
Project scientist(s): Woodhouse (UA, CLIMAS), McAfee (UNR), Harding (Lynker), Lehner (NCAR), Gutmann (NCAR), Wood (NCAR)
Summary: See project overview
Product(s):
Visit the Western Water Assessment website to access the extensive set of supporting materials including executive summary, individual chapters, webinars, and contact info.
Sub-seasonal to Seasonal Climate Forecasts for Water Managers
Institution(s): National Center for Atmospheric Research (NCAR), University of Colorado Boulder (CU)
Years: 2016 – 2019
PI(s): Wood (NCAR), Rajagopalan (CU)
Project scientist(s): Baker (CU)
Summary: NOAA’s Climate Test Bed program seeks to accelerate the transition of research advances in climate modeling products into operational predictions and projections. In order for the advances to inform water management, output from these products must be post-processed to enhance skill and translated into relevant spatial and temporal scales. Reclamation contributed to this effort through a project that developed real-time, watershed-based subseasonal-to-seasonal (S2S; two weeks to two months) climate forecasts. The prototype forecasts and associated skill analyses were created using raw data from two products – Climate Forecast System version 2 (CFSv2) and National Multi-Model Ensemble (NMME) – and are being disseminated on a public website to facilitate product development through interactions with water managers. Partial Least Squares Regression (PLSR) was used to post-process raw CFSv2 bi-weekly temperature and precipitation forecasts. To demonstrate an operational application of the products, an experimental streamflow forecast for the Colorado River Basin was developed using NMME watershed scale 1- and 3-month forecasts to weight Ensemble Streamflow Prediction (ESP) traces with a k-Nearest Neighbors algorithm. The results of comparing the experimental product to traditional forecasts via the Colorado River Streamflow Forecast Testbed indicate that S2S climate forecasts can potentially improve streamflow forecast skill. This project was jointly funded by NOAA.
Product(s):
Website with operational forecasts: http://hydro.rap.ucar.edu/s2s/
Additional journal articles in production
Using Multiobjective Evolutionary Algorithm (MOEA)-assisted Optimization to Explore Lake Mead Operating Policies
Institution(s): University of Colorado Boulder, Center for Advanced Decision Support for Water and Environmental Systems (CADSWES)
Years: 2016 – 2018
PI(s): Kasprzyk, Zagona
Project scientist(s): Alexander
Summary: In managing the Colorado River system, Reclamation must balance multiple performance objectives. Thoroughly exploring different ways of operating the system can provide information about the relationships between objectives that improves system understanding and supports dialogue. This study combined a Multiobjective Evolutionary Algorithm (MOEA) with the Colorado River Simulation (CRSS) to intelligently and efficiently devise and test thousands of different Lake Mead operating policies, searching for policies that provide the best tradeoffs between a set of conflicting objectives. The objectives account for different types of performance, e.g. maintaining Lake Mead and Lake Powell pool elevations vs. minimizing delivery reductions. The search does not prioritize any single objective, but allows users to apply their own preferences to a large set of policies after the optimization process is complete. This study’s successful use of MOEA-assisted optimization to explore a wide range of Lake Mead operations demonstrates its potential for integration into future planning studies.
Product(s):
Journal article (in production)
Colorado River Streamflow Forecast Testbed
Institution(s): University of Colorado Boulder (CU), Center for Advanced Decision Support for Water and Environmental Systems (CADSWES), National Center for Atmospheric Research (NCAR)
Years: 2016 – 2018
PI(s): Zagona (CU, CADSWES), Wood (NCAR), Rajagopalan (CU, CADSWES)
Project scientist(s): Baker
Summary: The Colorado Basin Streamflow Forecast Testbed (testbed) is a framework for assessing the performance of existing and potential future streamflow forecasts and operational projections in the Colorado River Basin. The ability to systematically and objectively evaluate forecasts will enable Reclamation to determine which forecasts offer the best performance in its one- to five-year-out projections. Forecasts are evaluated for hydrologic skill by analyzing annual Lake Powell unregulated inflow. Operational projections that result from running the streamflow forecasts through the Mid-Term probabilistic Operations Model (MTOM) are evaluated by analyzing projections of pool elevation, storage, and operating tiers at Lake Powell and Lake Mead. To date, the testbed has been used to evaluate the hydrologic and operational skill of the Colorado Basin River Forecast Center (CBRFC) most probable forecast, the full suite of Ensemble Streamflow Prediction (ESP) projections, and the median ESP trace, all of which are currently used in Reclamation modeling studies. The testbed is ready to accept experimental forecasts.
Product(s):
Streamflow forecast testbed
Final report (in review)
Journal article (in production)
For additional information, please contact us via email at: CRB-R2O@usbr.gov
Last updated: 2020-06-11
