Lauren E. Hay

Framework for Understanding Structural Errors (FUSE): A modular framework to diagnose differences between hydrological models

[1]xa0The problems of identifying the most appropriate model structure for a given problem and quantifying the uncertainty in model structure remain outstanding research challenges for the discipline of hydrology. Progress on these problems requires understanding of the nature of differences between models. This paper presents a methodology to diagnose differences in hydrological model structures: the Framework for Understanding Structural Errors (FUSE). FUSE was used to construct 79 unique model structures by combining components of 4 existing hydrological models. These new models were used to simulate streamflow in two of the basins used in the Model Parameter Estimation Experiment (MOPEX): the Guadalupe River (Texas) and the French Broad River (North Carolina). Results show that the new models produced simulations of streamflow that were at least as good as the simulations produced by the models that participated in the MOPEX experiment. Our initial application of the FUSE method for the Guadalupe River exposed relationships between model structure and model performance, suggesting that the choice of model structure is just as important as the choice of model parameters. However, further work is needed to evaluate model simulations using multiple criteria to diagnose the relative importance of model structural differences in various climate regimes and to assess the amount of independent information in each of the models. This work will be crucial to both identifying the most appropriate model structure for a given problem and quantifying the uncertainty in model structure. To facilitate research on these problems, the FORTRAN-90 source code for FUSE is available upon request from the lead author.

Hydrological responses to dynamically and statistically downscaled climate model output

Daily rainfall and surface temperature series were simulated for the Animas River basin, Colorado using dynamically and statistically downscaled output from the National Center for Environmental Prediction/ National Center for Atmospheric Research (NCEP/NCAR) re-analysis. A distributed hydrological model was then applied to the downscaled data. Relative to raw NCEP output, downscaled climate variables provided more realistic simulations of basin scale hydrology. However, the results highlight the sensitivity of modeled processes to the choice of downscaling technique, and point to the need for caution when interpreting future hydrological scenarios.

Simulation of precipitation by weather type analysis

A method of precipitation simulation that incorporates climatological information has been developed. A Markovian-based model is used to generate temporal sequences of six daily weather types: high pressure; coastal return; maritime tropical return; frontal maritime tropical return; cold frontal overrunning; and warm frontal overrunning. Precipitation values are assigned to individual days by using observed statistical relations between weather types and precipitation characteristics. When this method was applied to an area in the Delaware River basin, the statistics describing average precipitation, extreme precipitation, and drought conditions for simulated precipitation closely matched those of the observed data. Potential applications of this weather type precipitation model include climatic change research and modeling of temperature and evapotranspiration.

Rain-on-Snow Events in the Western United States

Rain-on-snow events pose a significant flood hazard in the western United States. This study provides a description of the spatial and temporal variability of the frequency of rain-on-snow events for 4318 sites in the western United States during water years (October through September) 1949–2003. Rain-on-snow events are found to be most common during the months of October through May; however, at sites in the interior western United States, rain-on-snow events can occur in substantial numbers as late as June and as early as September. An examination of the temporal variability of October through May rain-on-snow events indicates a mixture of increasing and decreasing trends in rain-on-snow events across the western United States. Decreasing trends in rain-on-snow events are most pronounced at lower elevations and are associated with trends toward fewer snowfall days and fewer precipitation days with snow on the ground. Rain-on-snow events are more (less) frequent in the northwestern (southwestern) United ...

A resampling procedure for generating conditioned daily weather sequences

[1]xa0A method is introduced to generate conditioned daily precipitation and temperature time series at multiple stations. The method resamples data from the historical record “nens” times for the period of interest (nens = number of ensemble members) and reorders the ensemble members to reconstruct the observed spatial (intersite) and temporal correlation statistics. The weather generator model is applied to 2307 stations in the contiguous United States and is shown to reproduce the observed spatial correlation between neighboring stations, the observed correlation between variables (e.g., between precipitation and temperature), and the observed temporal correlation between subsequent days in the generated weather sequence. The weather generator model is extended to produce sequences of weather that are conditioned on climate indices (in this case the Nino 3.4 index). Example illustrations of conditioned weather sequences are provided for a station in Arizona (Petrified Forest, 34.8°N, 109.9°W), where El Nino and La Nina conditions have a strong effect on winter precipitation. The conditioned weather sequences generated using the methods described in this paper are appropriate for use as input to hydrologic models to produce multiseason forecasts of streamflow.