Land Use And Cover Change Impacts On The Near-surface Wind Speed Over China

Based on the synthetic analysis and Empirical Orthogonal Function (EOF), the spatio-temporal characteristics of wind speed on different levels in resent30years over China and the impact of the pressure gradient force (PGF) and East-Asian Monsoon (EAM) on near-surface wind speed (SWS) and wind speed on925hPa,850hPa are investigated. Furthermore, based on the frictional wind model, the observation minus reanalysis (OMR) method, and the observation minus downscaling (OMD) method, the impact of the land use and cover change (LUCC) on SWS were analyzed. The results indicated that:A decrease in wind speed is detected in the data from observations dataset with the rates of decrease is-0.10m/(s·10yr) and-0.12m/(s·10yr) throughout China and East China Plain (ECP) region, respectively. The wind speed had a significant interannual and interdecadal change on the925hPa and850hPa in resent30years, the interannual and interdecadal change of the SWS is weaker than that of the wind speed on925hPa and850hPa. Meanwhile, the effect of the PGF to the wind speed is different on different levels, the impact of the PGF on the SWS is indistinctive on the near-surface, but the correlation between PGF and wind speed is distinct on925hPa and850hPa, the correlation coefficient is0.89and0.90, respectively, so the PGF is the main factor of the wind speed change on925hPa and850hPa. The long-term decrease of the SWS is not induced by the variability of the EAM too, the difference of SWS between strong EAM year and weak EAM year is not evident, the difference of SWS between strong East-Asian summer monsoon (EASM) and weak EASM year is0.017m/s, and the difference of SWS between strong East-Asian winter monsoon (EAWM) and weak EAWM year is-0.008m/s. Therefore, the downward trend of the SWS is not induced by the EAM and PGF over the ECP.The frictional wind model is reliable over ECP region, which could capture the main characteristics of the large-scale motion, such as the quasi-geostrophic balance on850hPa, meanwhile, the wind speeed which computed by the model (MWS) on1000hPa is consistent with that of reanalysis wind speed well with the probability of the extremums appearing at the same point in the two data serials (PEST) is95%during the study period, the averaged difference between the two serials is only-0.018m/s. In addition, based on the frictional wind model, the impact of the LUCC on the SWS is estimated, the averaged model wind speed is higher than that of observed SWS for0.5m/s during the study period. The drag coefficient (DC) has an evident increase trend, and that the positive correlation between DC and urbanization rate can be found, the correlation coefficient is0.88. Additionally, the difference between observed SWS and model wind speed (SWSD) could be regared as standard of the impact of LUCC on SWS. According to the linear regression, each0.24m/s decrease of the observed wind speed can be induced by10%rise of the urbanization rate over ECP region.Comparing with the ERA-Interim10m reanalysis wind speed, the downscaling wind speed model the mean value of the SWS and the interannual and seasonal variation of SWS is better. The statistical downscaling models are estimated through multiple stepwise regressions (MSR) of predictor principal components (PCs), which is stabilized and reliable. The multiple correlation coefficient of the MSR is more than0.7, and the standard deviation of the error is less than0.13m/s. the Brier Skill Score (BSS) is more than70%. Furthermore, based on the OMD mothod, the results show that the downward trend of the SWS which induced by LUCC is-0.18m/(s10yr), and that every ten percents increase of urbanization rate can lead to0.18m/s decrease of observed wind speed, which is smaller than the results of the frictional wind model and larger than the results of the OMR method.OMR method could be used to analyse the impact of the LUCC on the SWS. The influence of the LUCC is distinct, which may accord for0.71m/s wind speed decrease for the mean of the whole China,0.84m/s, and0.52m/s for large and small cities. However, the decrease of the SWS is higher over Beijing-Tianjin-Tangshan region (BTTR), the Yangtze River Delta region (YRDR), and Pearl River Delta region (PRDR) than that over the whole China, therefore, the decrease of the SWS increase with the LUCC increase, the positive correlation between urbanization and the difference between reanalysis wind speed and observed wind speed is distinct, correlation coefficient is0.86. According to the linear relation, every ten percents increase of urbanization rate can lead to0.11m/s decrease of observed wind speed, which is smaller than the influence of the LUCC on SWS over ECP region.