Rojas-González, Alejandra M.

Profile Picture

Filters

20121
Reset filters

Settings

Citations

Publication Search Results

Now showing 1 - 1 of 1
  • Thumbnail Image
    Publication
    Flood prediction limitations in small watersheds with mountains terrain and high rainfall variability
    (2012) Rojas-González, Alejandra M.; Harmsen, Eric W.; College of Engineering; Zapata López, Raúl E.; Rivera Santos, Jorge; Segarra García, Rafael; Department of Civil Engineering; Mercado Irizarry, Aurelio
    An evaluation of the interrelation between different up-scaling parameters and inputs were evaluated to quantify their influence on hydrologic predictability in complex terrain and small watersheds. An up-scaling experiment was performed, consisting of increasing the grid size to produce incrementally coarser resolution maps of each parameter, terrain and rainfall inputs. Each resolution was evaluated by an ensemble approach and generalized likelihood uncertainty estimation (GLUE) methodology using high resolution rain gauge network (rainfall resolution of 100 m) and fully distributed hydrologic model (10 meters). Each parameter perturbation, hydrologic model resolution, and rainfall resolution combination were modeled producing deterministic forecasts called “ensemble members”. Objective functions were used to evaluate the behavior of each ensemble with observed data using the variables time to peak, runoff depth and peak flow observations. Ensemble skill was evaluated using scalar measures of accuracy for continuous prediction as mean absolute errors (MAE), root mean square error (RMSE) and bias between the average ensembles to observation variable. Probabilistic distribution functions (PDF) were generated for each ensemble and prediction skill was measured by ranked probability score (RPS). Based on the analyses presented in this research, the recommended upscaled rainfall resolution, which will provide equivalent accuracy with the 100 m rainfall resolution, is 1000 m, and the recommended upscaled hydrologic model grid resolution, which will provide equivalent accuracy with the 10 m resolution, is 100 m.