Hydrologic Modeling

NASA-CASA Project

Hydrological Routing Algorithm

SWAT

Related Links

Hydrological Routing Algorithm (HYDRA)

HYDRA Simulations in California

We are currently using HYDRA to model the state of California in order to understand the effects of land cover and climate change on water resources. HYDRA's input data for this project is at 1-km resolution. In the map showing early results (see below), the black lines delineate California's major waterways and the grey, white, blue and orange regions represent HYDRA's estimation of the location of water channels and the corresponding water flow rate. (Grey identifies the lowest water flow and orange shows the highest.) To assess HYDRA's ability to estimate actual water flows, we have compared HYDRA's results with gauge station data throughout the state. This comparison yielded 44 points where HYDRA's estimate of the yearly water volume is within +/-30% of the actual yearly water volume (referred to as "best-fit" points). Out of those 44 best-fit points, four example points are described in more detail below.

At the Ft. Seward gauge station on the Eel River, HYDRA's estimate of the yearly water volume is just 19% lower than the actual yearly water volume. The river's monthly variation is represented in Figure 2. The pronounced seasonal curve, reflecting much wetter winters and relatively dry summers, is typical of the best-fit pixels in the Northern Coastal Range. Moreover, the Eel River is contained within a fairly complex watershed. Under these circumstances, HYDRA's estimate of actual water flow improves as the model progresses downstream within the watershed. Our comparison includes seven gauge stations that are on the Eel River watershed, four of which are best-fits. Other complex watersheds that display similar characteristics include the Klamath and the San Joaquin.

Similar to the Eel River, the monthly distribution of the water flow volumes of the Navarro River (Figure 3) is also characteristic of the best-fit pixels in the Northern Coastal Range. In contrast to the Eel River's complex watershed, however, the Navarro River is a discrete water system. It is a relatively small river and it has no major tributaries. HYDRA frequently provides good estimates on rivers with this characteristic, including Redwood Creek and the Mad River. HYDRA's estimate of the yearly water volume on the Navarro River (at Navarro, CA) is 24% larger than the actual yearly water volume.

HYDRA matches the yearly water volume for the American River gauge (below Auburn Dam) with a small (7%) error. This gauge is one of ten points in the Sierras that are best-fits, which suggests that HYDRA performs well in this region. One contributing factor is that the topography is varied enough in this region that, at 1-km resolution, HYDRA is able to accurately estimate the actual river channels. In Figure 4, the monthly distribution of the actual flow volumes constitutes a somewhat flatter curve than HYDRA's estimates. This result is reasonable since dam water that is held back in the wet months is discharged during the dry summer months. Incidentally, good HYDRA results also occur in the vicinity of dams along the Pit River and the San Joaquin River.

The Mojave River, like the Navarro River, is a discrete water system that HYDRA is able to accurately estimate. At the gauge station near Hodge, CA, for example, HYDRA overestimates the yearly water volume by just 15%, and Figure 5 illustrates a good correlation of the monthly totals. This data point was included to demonstrate that HYDRA can perform well in a dryer climate. This observation is further evidenced by the seven best-fit pixels that occur in the vicinity of Los Angeles and San Diego.

For further information regarding the CASA-HYDRA model application, contact Christopher Potter at cpotter@mail.arc.nasa.gov or see NASA Ames web site at http://geo.arc.nasa.gov/sge/casa

Responsible Official: Christopher Potter
Website Design: Cynthia Lim