Analyzing the influence of large wind farms over rainfall pattern using radar data
Abstract
Watershed run-off models were mostly developed using point rainfall observations in the past. Water quality model use watershed run-off models. The water quality modeling can be done in a better manner if radar rainfall data is used in the modeling process. However, radar rainfall is not directly available for this purpose. Radar data is available through public domain as radar signals, which are converted to rainfall rates using a Z-R relationship (Pathak and Teegavarapu (2010)). Rendon et al. (2011) proposed a new approach using the mean field bias to derive radar specific Z-R relationship for hydrologic operations. In this study, a neural network approach is attempted to find the daily rainfall magnitude from radar signals. Daily observations from 109 rain gauge stations were used in this analysis. Literature reported ZR parameters "a" and "b" values for stratiform storm was used initially to calculate the rainfall rate. Subsequently, "a" and "b" values were revised using the procedure recommend by Rendon et al. (2011). Two more models were developed using a neural network. The results of conventional ZR using literature reported "a" and "b" values, modified ZR method proposed by Rendon et al. (2011), the fuzzy ANN model and the ANN model results were compared using indices: mean square error and mean related error. ANN models performed better. Large wind farms are being built in the feasible locations of the United States as this energy production is replenishable and environmentally friendly. This research examines the changes in the precipitation pattern due to the large wind farms using a case study. Large wind farms were built recently in the Benton, White, and Warren Counties of Indiana. This region was considered for this case study. NEXRAD radar rainfall data was used in this analysis. Storms that occurred before and after wind farms installations were compared using radar data to identify the changes in precipitation patterns. Chicago NEXRAD radar data, which covers the study region, was used. To understand the changes in precipitation pattern caused by wind farm, two adjacent regional windows involving three counties each were considered. The results indicate few changes in the rainfall pattern in this region.
Degree
M.S.E.
Advisors
Viswanathan, Purdue University.
Subject Area
Civil engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.