• Home
• Resources
•       Publications
•       HSPF Support
•       Workshops
•       Downloads & Software
• Services
•       Watershed Model Applications
•       Software Development
•       Watershed Model Research
• Projects
• About Us
•       Who We Are
•       What We Do
• Careers
• Contact Us

Clients: U.S. EPA Office of Research and Development, Washington, DC

Climatic variability leads to changes in water resources and aquatic ecosystems. Individuals and groups involved in watershed management understand this from a conceptual perspective. However, in many cases they lack the ability to assess the impact of change on their area of interest using a tool that is meaningful to them at a watershed scale. The U.S. EPA ORD Global Change Program has funded a project that provided a tool to assist both in the analysis of impacts of projected climate change on specific hydrologic and water quality indices familiar to watershed managers and stakeholders, and analyzes the magnitude of climate change required to make a specific change in a streamflow or water quality endpoint.

The Climate Assessment Tool (CAT) has been implemented as a component of the development effort that created Version 4.0 of the U.S. EPA Office of Water's BASINS modeling system. CAT allows BASINS users to simulate or modify weather time series and use these data with the Hydrological Simulation Program - FORTRAN (HSPF) watershed model. In addition to the HSPF model, BASINS includes an extensive suite of data and analytical tools selected for their applicability to the topic of watershed assessment and analysis; these tools and data have provided a firm foundation for the Climate Assessment Tool. The software design of BASINS is based on a set of interoperable components. Each component performs specific tasks such as data download, time series visualization or map projection. To support the Climate Assessment Tool development effort, new tools have been integrated into BASINS that allow the user to generate a weather scenario for use in an assessment and to adjust a weather scenario based on modeling result endpoints. Additionally, existing BASINS tools used to summarize and compare modeling results have been refactored to provide additional capabilities and flexibility.

Figure 1: Example of plot that can be generated using CAT results.

Water and watershed systems are influenced by the amount, form, seasonality, and event characteristics of precipitation, as well as temperature, solar radiation and wind that affect evaporative loss. Ultimately, these changes may be reflected in key management targets, such as TMDLs, duration flow events (e.g., 7Q10), maximum water temperatures, or nutrient loads. System vulnerability, or susceptibility to harm, is determined by sensitivity together with socio-economic factors influencing the ability to adapt to new conditions. An understanding of system sensitivity is thus a necessary foundation for conducting watershed-scale vulnerability analyses.

A range of plausible future climatic conditions and events can be estimated in most areas based on records of historical and paleo-extreme events, scenario based model projections, and consideration of hypothetical climatic events such as re-combinations of historical events (e.g. occurrence of an extreme dry year for 10 consecutive years). Given this information, sensitivity analysis can provide managers with an understanding of the plausible range of impacts, and help guide the development of strategies and practices to reduce the likelihood of future harm.

Figure 2: Contour plot showing the response of mean annual nitrogen loading (pounds*1000 per year) in the Monocacy River to changes in mean annual temperature and precipitation based on the analysis using synthetic scenarios.

EPA's Chesapeake Bay Program, in partnership with other government agencies, is working to restore water quality and living resources in the Chesapeake Bay Watershed and in the Bay itself. Water quality modeling plays a central role in planning and prioritizing regulatory programs and restoration activities. The potential impacts of climate change on the Bay watershed are not well understood, but are likely to be significant. Changes in climate may also alter the efficiency of key nutrient reduction strategies (i.e., Best Management Practices) proposed by the Bay Program, but have not been addressed in previous Bay modeling efforts.

An initial case study was conducted with collaboration between the U.S. EPA ORD Global Climate Research Program, the U.S. EPA Chesapeake Bay Program Office (CBPO) and the U.S. EPA Office of Water. CAT/HSPF was applied to a tributary-scale watershed (Monocacy River Watershed) within the Chesapeake Watershed. Climate change scenarios were created by modifying hourly temperature and precipitation records for a 16-year period of historical data, from 1984 to 2000, from eight NCDC weather stations located within or near the watershed. The three ways the data were changed are: 1) as a constant multiplier applied equally to all events within the specified season, 2) as a constant multiplier applied only to the largest 30% of events within the specified season, and 3) as a constant multiplier applied only to the largest10% of events within the specified season. Sensitivity analyses were developed and modeled that evaluated the changes in relevant hydrologic and water quality endpoints that resulted from a range of these plausible changes. As the study was conducted to show an example of the types of analysis supported by BASINS CAT, the results were only given for a single endpoint, mean annual nitrogen loading. The results are shown in figures 2 and 3.

Demonstrating the Climate Assessment Tool in this context enables the leveraging of several decades of detailed HSPF watershed model development by the CBPO. At the same time there is an opportunity for the CBPO to benefit from a Climate Assessment Tool application that is focused on hydrologic and water quality endpoints and thresholds that are relevant to land use and Best Management Practices (BMPs) of greatest interest/concern in the Mid-Atlantic area.

Figure 3: Scatter plot showing the response of mean annual nitrogen loading (indicated by the color scale in pounds* 1000 per year) in the Monocacy River to model based climate change scenarios.

*Circles represent projection based on the A2 emissions storyline, triangles represent projections based on the B2 emissions storyline, and the star represents current mean annual temperature and precipitation

More recently, projected land use and management practice changes were combined with the climate change scenarios to explore the impacts of future development. BMPs proposed for the Monocacy watershed were also modeled, showing effectiveness in all combinations of future climate and landuse change. Using CAT jointly with the capabilities of HSPF allows the ability to assess the combined impacts of climate change, land use change, implementation of BMPs and other watershed management concerns.

Future landuse change scenarios were developed based on projected changes in housing density from EPA's Integrated Climate and Landuse Scenarios (ICLUS) dataset. ICLUS data are derived from a demographic model and a spatial allocation model that distributes the population projections generated by the demographic model into housing units across the landscape. Five ICLUS datasets were obtained to depict current (2000) and future (2030 A2, 2030 B2, 2090 A2, 2090 B2) landuse scenarios that were modeled. Given the complexities of the numerous scenario combinations in this study, detailed modeling of all proposed BMPs in the Monocacy was not feasible. Rather, the goal of this study was to model a simplified BMP implementation to assess the watershed's sensitivity to BMPs and their combined effects with climate and landuse change. Though simplified in their model implementation, a number of reports related to the Monocacy watershed and the entire Chesapeake Bay were used as resources in developing the BMPs. With the wide array of plausible scenarios investigated in this study, summarizing results is especially challenging. For the sake of this description, a brief summary of findings across the three main variable spaces (climate model, landuse, BMP) is provided. For this discussion, Figure 4 shows sample results for various scenarios, with current mean values noted by the solid vertical line and the range of endpoint values across the seven climate models depicted by the horizontal bars.

Figure 4:Total N Load projections across 7 climate models for 2030 and 2090. Data legend: Emission storylines (A2, B2), Precipitation Intensity levels (>10%, >30%, and All), landuse (Future and Current), and BMP implementation (Yes, No).