Comparison of deterministic wind speed generation models for wind erosion modeling

Abstract

The diurnal patterns of hourly wind speed data are fundamentally important in precise wind erosion modeling. Several deterministic wind speed simulation models, including the Guo (1/π), Guo (1/2), wind generator (WINDGEN), Ephrath, Debele, and Goudriaan models, have been developed to quantitatively generate hourly wind speed data. In this study, the performances of these popular deterministic wind speed generation models were investigated using observed hourly wind speed data obtained from 191 climate stations in arid and semi-arid areas of China. The performances of the Guo (1/π), Guo (1/2) and WINDGEN models were the best for hourly wind speed generation. Linear regression analysis revealed that all models tended to underestimate wind speeds. The patterns of diurnal variation and the amplitude of the measured hourly wind speed data generally governed the model performance. The performances of the Guo (1/π), WINDGEN, Guo (1/2), Goudriaan, Debele, and Ephrath models successively decreased for the daily average erosive wind power density (AWPD) estimation. To improve the Guo (1/π) and Guo (1/2) models for AWPD evaluation, the calibrated parameters (π and 2) of the second term in the Guo (1/π) and Guo (1/2) models showed different values for each station. The accuracy of the AWPD obtained by the Guo (1/π) and Guo (1/2) models with the calibrated parameters were significantly improved, which could further influence the accuracy of wind erosion modeling. More studies are required to investigate the relationship between the calibrated parameters in the Guo (1/π) and Guo (1/2) models and geographic conditions linked with wind erosion.