The 2010 Resources Planning Act (RPA) Water Assessment evaluates the vulnerability of the United States water supply to shortage. The RPA Assessment is produced every 10 years in response to the Renewable Resources Planning Act of 1974. Reports from the 2010 Assessment are now being released.
Although aggregate water withdrawal in the United States has leveled off in recent years and water use efficiency has been improving, future population and income growth is likely to place additional demands on raw water supplies. In addition, climatic change is increasing hydrologic uncertainty. Taken together, these forces are making careful water management ever more important and call for a realistic broad-scale understanding of the vulnerability of our water supply to shortage.
The 2010 RPA Water Assessment estimates the vulnerability of renewable water supply to shortage in the contiguous 48 states of the U.S. over the 21st century in light of projected socioeconomic and climatic changes. To capture in a rough sense the uncertainty about the estimates of vulnerability, the probability of shortage is estimated for three alternative scenarios of future socio-economic conditions (corresponding to the IPCC’s SRES A1B, A2, and B2 emission scenarios). Each scenario is modeled using downscaled estimates of weather variables from three different global climate models.
Our approach to estimating vulnerability begins with projections of socioeconomic conditions (including population, per-capita income, irrigated area, and other basic drivers of resource use), which affect long term climatic conditions as well as water demand and water management. Climate in turn affects both water yield (via temperature, precipitation, and other weather variables) and water demand (via effects on agricultural and landscape irrigation and water use at thermoelectric plants).
Water supply is determined by the management, via reservoir storage and diversion, of available water yield. Water demand in this framework is desired consumptive water use, the net amount of water depletion that would occur if water supply were no more limiting to future water use than it has been to recent levels of water use. Vulnerability is estimated by comparing supply and demand. This approach to vulnerability isolates those areas that are most likely to face water supply challenges.
Vulnerability is estimated for 98 basins, called Assessment Sub-Regions (ASRs), which make up the contiguous US. The ASRs are nearly identical to those defined by the US Water Resources Council for its second national water assessment. A thorough characterization of water supply must account for the natural and man-made water networks that redistribute water on the surface. In this framework, two or more ASRs are part of the same network when a sequence of water links, either natural (due to natural upstream to downstream flow) or artificial (via water diversions), connects them. The ASR-based water supply system for the US consists of three multi-ASR networks and 15 single-ASR systems. The biggest of the three multi-ASR networks includes 69 ASRs in the central and western US. The other two networks include, respectively, 10 ASRs in the Northeast and four ASRs in the Southeast. Of the 15 single-ASR systems, eight drain to the ocean, five drain into Canada, and two are closed basins.
A hydrologic network model was used to simulate water management in each water network. The model performs year-by-year linear optimizations of water allocation in a network consisting of a system of nodes connected by links. Each link is subject to capacity constraints and is assigned a priority that reflects the operating rules of the system. Each node is a point of water storage, reservoir evaporation, and/or water diversion. The simulations provide annual values of water flows in any link, storage levels and reservoir evaporation in each ASR, and water assigned to each demand, all of which depend both on climate and the set of priorities.
Ideally, the priorities would represent all of the detailed agreements about water storage and allocation that exist across the country. Lacking information on many of these agreements, we implemented a simple set of priorities in the following order: (1) in-stream flow requirements, (2) trans-ASR diversions, (3) consumptive water uses, and (4) reservoir storage. These priorities recognize the importance of guaranteeing a minimal amount of water for environmental and ecosystem needs before water is diverted for other uses, and allows trans-basin diversions to occur before within-basin diversions. For multi-ASR networks, water demands belonging to the same category were assigned the same priority regardless of their position in the network. Because storage was assigned the lowest priority level, water is stored in a given year only after all the demands reachable by that reservoir are satisfied. Water stored at the end of one year, minus an evaporation loss, is available for use the next year.
Using the modeling framework of the 2010 Water Assessment, we will examine the effect on projected water supply shortages of a set of adaptation measures including additional reservoir shortage, new trans-basin diversions, and reductions in the drivers of water demand.
Changes from the 2010 assessment will include: (1) a finer spatial scale (the 204 4-digit basins of the contiguous U.S.), (2) a finer temporal scale (monthly), (3) a different water yield model (VIC), (4) updated estimates of demand for water in the agricultural and thermoelectric sectors, and (5) a different set of scenarios and GCMs. The downscaled estimates of climate variables are taken from the CMIP5 (Coupled Model Intercomparison Project Phase 5) website. The seven GCMs and the related scenarios are as follows:
|CSIRO Mk3.6.0||RCP 26, 45, and 85|
|BCC-CSM1.1||RCP 26, 45, 60, and 85|
|CanESM2||RCP 26, 45, and 85|
|GFDL-ESM2M||RCP 45, 60, and 85|
|IPSL-CM5A-LR||RCP 26, 45, 60, and 85|
|MIROC-ESM||RCP 26, 45, 60, and 85|
|MPI-ESM-LR||RCP 26, 45, and 85|
The RCP (representative concentration pathways) scenarios represent four possible greenhouse gas concentration trajectories adopted by the IPCC for its Fifth Assessment Report. They correspond to increases in global radiative forcing, from preindustrial times to 2100, of 2.6, 4.5, 6.0, and 8.5 W/m2. Data for many GCMs are available at the CMIP5 site. Among the criteria we used for selection of the seven GCMs we chose is that they provide data needed for implementing the VIC hydrologic model.
Employing the enhanced framework summarized above, we will include a detailed analysis of water demands obtained using a CGE (computable general equilibrium) framework for the Upper Colorado River Basin. This study will allow a more nuanced projection of water supply shortages and analysis of adaptation options.