Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Chine → 2021 → Simulation of Crop Soil Water Availability and Drought Assessment in the Northen Songnen Plain

Harbin Normal University (2021)

Simulation of Crop Soil Water Availability and Drought Assessment in the Northen Songnen Plain


Titre : Simulation of Crop Soil Water Availability and Drought Assessment in the Northen Songnen Plain

Auteur : 王勇

Grade : Doctoral Dissertation 2021

Université : Harbin Normal University

Résumé partiel
Drought is the most important meteorological disaster in China’s agriculture.It is of great significance to accurately evaluate the temporal and spatial distribution and severity of crop drought for the improvement of crop growth and grain yield.At present,domestic and foreign scholars have put forward a variety of drought indexes for the evaluation of agricultural drought,which can be divided into two categories : drought index based on meteorological data and drought index based on remote sensing monitoring data.However,the current agricultural drought index is still limited to the station precipitation,vegetation index,or single soil moisture data.In view of this,this study systematically analyzed the threshold range of soil water availability under different crop types,different growth periods,different soil texture,precipitation,and climate conditions in the northern Songnen Plain which was the main grain-producing area of China and one of three black soil regions of the world.On this basis,combined with the surface and different depth of soil moisture data,the agricultural drought index which was called the Crop Loss and Surplus Water(CLSW)was put forward,the severity of crop drought in the study area was quantitatively analyzed in the spatial and temporal distribution,and the quantitative evaluation model of regional crop drought was established.The main research contents and conclusions are as follows :(1)Study on scaling method of surface soil moisture data.Based on the Random Forest(RF)regression algorithm and MODIS optical remote sensing data,the SMAP microwave remote sensing data were converted to scale.The scale conversion process and method for microwave remote sensing surface soil moisture data were constructed,and the surface soil moisture data with high spatial and temporal resolution were obtained.The results show that this method can not only obtained high spatial-temporal resolution surface soil moisture data but also improved the accuracy of surface soil moisture data.(2)Research on crop classification method based on multi feature parameter set.Based on the traditional time series vegetation index and the parameter set of crop phenological characteristics,a new parameter set of crop classification characteristics was constructed by adding time series surface soil moisture data.Based on Random Forest(RF)classification algorithm,the traditional and new feature parameter sets were used as input parameters of the model,the crop planting areas were extracted.The results show that the new set of characteristic parameters of crop classification can improve the classification accuracy of crops,especially,the precision of rice planting area extraction was improved.(3)Retrieval method of deep soil moisture based on data assimilation.According to the land surface data assimilation framework,a data assimilation system based on Ensemble Kalman filter(En KF)and community land model(CLM)was constructed

Mots clés : soil moisture ;downscaling;crop classification ;land surface processes and data assimilation ;soil water movement ;drought index ;

Présentation (CNKI)

Page publiée le 23 octobre 2021