| Wind erosion over cropland is not only a concern for sustainable development of agriculture,but also affects air quality,climate and human health.The overarching goal of this study is to quantify global emissions from wind erosion using a state-of-the-art dust emission model FENGSHA,coupled with dynamic source map and global weather model.First,we determine the global source map of cropland wind erosion by satellite remote sensing data products.Next,we improve the FEGNSHA dust emission model and quantify global cropland dust emissions.We also investigate the similarities and differences of results obtained through FENGSHA and other commonly used dust schemes.Finally,we validated the newly developed cropland dust emissions by simulating the cropland dust events in northeastern China and central United States.The main conclusions of this study are as follows:?1?A new method was developed to calculate the source map of global cropland wind erosion by using the global cropland area data?15 crops?,crop calendar data,and remote sensing data.The estimated yearly total area of global cropland wind erosion is 2.59×1011 hectares.We also use the MODIS NDVI to calculate the global cropland soil wind erosion area.In general,the total global cropland area of wind erosion is relatively lower compare to the results from the cropland data method.The spatial distribution of satellite-derived source map is generally consistent with that from the crop calendar-based source map.?2?The global distribution of cropland dust emission was simulated using the FENGSHA dust emission module.The results showed that the global cropland wind erosion of cropland soil is highly sensitive to the source map of global cropland wind erosion.The total dust emission of cropland wind erosion is 1.40×109 g/s.Other seven dust schemes?SH2004?AG2001?MB1995?GP1988?OW1964?LL1978?KK2014?were also used to simulate the global dust emission of cropland wind erosion and the dust scheme KK2014 can only simulate the dust emission flux.The results showed that the dust emission of cropland wind erosion simulated by SH2004scheme is relatively high,with the total dust emission of 8.50×1012 g/s,which is generally underestimated comparing to the results of FENGSHA module.The results of other five dust schemes are lower than the results of FENGSHA module.The highest total dust emission of AG2001 scheme was 6.50×108 g/s.In addition,the results of FEGNSHA module and other six schemes showed that the dust emissions of cropland wind erosion were high in late winter and early spring,and lowest in early summer,which is positively correlated with the wind speed.?3?According to the existing particle size distribution methods,the distribution of particulate matter in the global emission inventory of cropland wind erosion was studied.The bin method was selected to effectively obtain the distribution of PM2.5and PM10,which is conducive to the study of model simulation.The results showed that the total PM2.5 and PM10 emissions of global cropland wind erosion from FEGNSHA module are 1.93×108 and 9.49×108 g/s,respectively.Among them,the annual total emissions of PM10 from farmland soil wind erosion are 4.9 times that of total PM2.5 emissions per year.This result is applicable to the other six dust schemes.?4?The FENGSHA and three other schemes?SHAO2004,MB1995 and LL1978?were used to simulate three cropland dust events,including the October 2012 dust storm in the Central Pains of the USA,and the May 2013 and April 2018 storms in Northeast China.It was found that the simulation results of the cropland wind erosion by FENGSHA agree better with observations during these cropland dust events.Although the distribution characteristics of hourly PM10 concentration are slightly deviated from the observations,the overall trend is consistent with them and the peak of the PM10 concentration is similar to the observations.Therefore,the spatial and temporal distributions simulated by FENGSHA are in good agreement with dust events.The average relative mean square error?RMSE?,average normalized standard deviation?NMB?and the average normalized mean bias?NME?in 2012 were 108.93?g/m2,-0.45?close to 0?and 91.68%,respectively.And the value in 2018 are 70.91?g/m2,-0.47 and 70.26%,respectively.It means that the simulation results of FENGSHA module for dust events in China are slightly better than those in the USA.However,the results of FENGSHA module also has deviation from the observation,due likely to soil texture and other surface conditions that affect the threshold friction velocity of cropland wind erosion. |
PDF Full Text Request