| The rapid and accurate acquisition of the field crop coefficient K_c is the key to estimating the daily evapotranspiration of crops in dryland.Based on the field corn under different water treatments in Zhaojun town Experimental Station in Dalate Qi,Inner Mongolia,2017-2018.This paper uses the self-developed UAV remote sensing platform to obtain the canopy spectral image of maize and adjust the rotational speed of the sprinkled irrigation to achieve different water treatment in each region.It simultaneously collect ground data to achieve different Co-monitoring of remote sensing and ground sensors.The crop coefficient of maize was calculated by the FAO-56 dual crop coefficient method calibrated by meteorological factors and crop coverage.The relationship between the crop coefficient and the four different types of vegetation indices were studied(ratio vegetation index SR,normalized difference vegetation index NDVI,soil adjusted vegetation index SAVI,enhanced vegetation index EVI).Meanwhile,the relationship between the crop coefficient and leaf area index(LAI),surface soil water content(SWC)was analyzed.The corn crop coefficients under different climatic conditions and different water stress conditions and feasibility of co-estimation of UAV remote sensing and ground sensors were analyzed.The main research contents and conclusions of this paper are as follows:(1)Correlation analysis between soil moisture content and crop coefficient at different depths showed that under different water stress treatments,surface soil moisture(30cm)was at a strong correlation level and the highest correlation was achieved 0.72(P<0.01)under the most severe water stress.Under heavy rainfall conditions,the correlation between shallow soil moisture content and crop coefficient is:10 cm SWC>20 cm SWC>30 cm SWC.(2)Under different water stress conditions,there was no significant difference between the leaf area index LAI and the crop coefficient.They were all at a high level(r=0.45~0.60,P<0.05).However,under strong rainfall conditions,the LAI correlation coefficient is low(r=0.04~0.27).(3)The correlation between the four different types of vegetation indices and crop coefficient under different water treatment and different climatic conditions was:SR>NDVI>EVI>SAVI.Moreover,as the degree of water stress is aggravated,the correlation gradually decreases.This is because the vegetation index has a certain hysteresis in the late growth stage,and the response to water stress is not high.(4)The 2017 K_c stepwise regression model based on the ratio vegetation index,leaf area index and surface soil moisture was established and verified.The coefficient of determination,root mean square error and normalized root mean square error were0.60,0.21 and 23.35%,respectively.It indicates that the established K_c estimation model have better estimation accuracy under different water stresses in the arid area.However,under heavy rainfall conditions in 2018,the model has been verified.The coefficient of determination,root mean square error and normalized root mean square error were 0.24,0.16 and 15.5%,respectively.It can be seen that the three variables of the model explain the crop coefficients is not high enough and the accuracy is not enough under heavy rainfall conditions,but the RMSE is reduced,indicating that the difference between the estimated amount and the estimated amount is relatively stable.(5)Based on the comparison between the actual evapotranspiration ET calculated by the FAO-56 dual crop coefficient method and the simulated evapotranspiration established by the model,the simulated values will gradually overestimate the actual evapotranspiration of the crop as the degree of water stress increases.However,under heavy rainfall conditions,although the mean difference between the two growth stages was not significant,the difference in daily evapotranspiration was significant,indicating that the model has low estimation accuracy under heavy rainfall conditions. |
PDF Full Text Request