| Title: | Deriving spatially-distributed precipitation using the artificial neural network and multi-linear regression models |
| Authors: | Sharma, S., S. Isik, P. Srivastava and L. Kalin |
| Year: | 2013 |
| Journal: | Journal of Hydrologic Engineering |
| Volume (Issue): | 18(2) |
| Pages: | 194–205 |
| Article ID: | |
| DOI: | 10.1061/(ASCE)HE.1943-5584.0000617 |
| URL (non-DOI journals): | |
| Model: | SWAT |
| Broad Application Category: | hydrologic only |
| Primary Application Category: | climate data effects |
| Secondary Application Category: | hydrologic assessment |
| Watershed Description: | 1,837 km^2 Saugahatchee Creek, located in southeast Alabama, U.S. |
| Calibration Summary: | |
| Validation Summary: | |
| General Comments: | |
| Abstract: | Precipitation is the primary driver for hydrologic modeling. Since hydrologic models often
require long term, spatially-distributed precipitation datasets for calibration and validation, a
novel approach was developed to generate spatially-distributed precipitation data using the
Artificial Neural Network (ANN) for the periods when NEXRAD data are either unavailable or
the quality of the NEXRAD data is not good. The Multi-Linear Regression (MLR) technique
was also evaluated for completeness. The study’s focus was the Saugahatchee Creek watershed
in southeast Alabama. In the study area, the wet seasons are dominated by frontal precipitations,
while the dry seasons primarily contain patchy, convective thunderstorms. The basic approach
was to train and validate the ANN and MLR models using recent NEXRAD and rain gauge
precipitations, and then use the trained model with the rain gauge precipitation data to generate
past, spatially-distributed precipitation estimates at the NEXRAD grid locations. Results showed
that, for the testing period, the ANN-simulated wet season precipitations in all the NEXRAD
grids had a Nash Sutcliff Efficiency (NSE) greater than or equal to 0.72 and a Mass Balance
Error (MBE) less than or equal to 14%. The same model performance parameters were 0.65 and
17% for the dry season. The MLR model did not perform as well as the ANN model. For the
MLR model, the wet season MBE ranged from 13 to 48%, while the dry season MBE ranged
from 0.1 to 36% on the testing dataset. An uncalibrated Soil and Water Assessment Tool
(SWAT) model was used to assess the improvements in stream flow simulations with the ANN-simulated
spatially-distributed precipitation data. It was found that the stream flow simulations
using ANN-generated, spatially-distributed precipitations were closer to the observed stream
flows as compared to stream flows generated using the rain gauge precipitations. Overall, the
results suggest that the method developed in this study can be used to generate past, spatially distributed
precipitations at NEXRAD grids locations. |
| Language: | English |
| Keywords: | Precipitation, NEXRAD, Artificial Neural Network, Multi-Linear Regression, Hydrologic Modeling, SWAT |