Based on hourly temperature and precipitation grid data of CMA land data assimilation system (CLDAS), ground meteorological observation data at 119 weather stations of Inner Mongolia and the spatial distribution of irrigated crop fields, the key meteorological factors and climatic division indexes were determined by considering fully planting technology and agricultural climatic characteristics of sunflower following spring wheat harvest in agro-climatic ecological regions of Inner Mongolia. And taking suitable growth days and climate suitability in the whole growth period as zoning indicators, the refined climatic suitability division of sunflower following spring wheat harvest in Inner Mongolia was studied under the operation of ArcGIS. The results show that the heat shortage and drought in growing season mainly restricted the regular growth and yield formation of sunflower following spring wheat harvest in Inner Mongolia, and the whole region was divided into unsuitable zone, more suitable zone, suitable zone and the most suitable zone. The irrigation conditions in Hetao irrigation district, Tumochuan plain and West Liaohe plain were better, which were particularly suitable to the growth of sunflower following spring wheat harvest, so these regions were the most suitable districts of sunflower growing. Due to serious shortage of heat resource in most areas of middle and eastern Inner Mongolia and the deficit of precipitation in the north of mid-western Inner Mongolia after the transplanting, the sunflower following spring wheat harvest was quite immature, so these regions were unsuitable areas of sunflower growing. The climatic suitability division of sunflower following spring wheat harvest based on grid data of CLDAS was consistent with the division distribution based on meteorological data of weather stations in Inner Mongolia, but the details of CLDAS data division were more refined than that of stations data. Hetao irrigation district was the most suitable area of sunflower following spring wheat harvest in Inner Mongolia, which was accord with the current layout of sunflower following spring wheat harvest. Therefore, this division result could provide some references for the rational layout of sunflower following spring wheat harvest in Inner Mongolia under the background of climate change.

 Based on snow depth during 1981-2010 from two reanalysis data sets :ERA-Interim/Land (hereafter referred to as Land) and ERA-20C (hereafter referred to as 20C), the observation data of meteorological stations in Inner Mongolia were taken as the standard, the applicability of snow depth of reanalysis in Inner Mongolia was evaluated comprehensively. The results were as follows: (1) Two reanalysis data sets both reproduced the spatial distribution of snow depth in Inner Mongolia, showing a significant decline trend from northeast to southwest, and 20C reanalysis described well in spatial distribution in descending order from northeast to southwest. (2) Monthly variation of average snow depth of Land reanalysis was compared with observations, it was found that a negative deviation existed on the whole and average snow depth of Land reanalysis was systematically underestimated. However, the simulation effect of 20C reanalysis in eastern and western region was good, and simulated snow depth of  20C reanalysis was compared with observations, it was found that the positive deviation of 20C reanalysis was larger in the central region. (3) The two reanalysis data sets reprensented well in monthly variation of snow depth in eastern, central and western Inner Mongolia. The simulation effect of 20C reanalysis was better than that of Land reanalysis. Simulation effect was best in the eastern region, followed by the central region, and the western region was worst. (4) The mean deviation and root mean square error of snow depth of two reanalysis data sets were relatively small, correlation coefficient between reanalysis and observed snow depth passd the 0.001 significance test. The difference of snow depth of two reanalysis data sets was smaller, but the difference between reanalysis and observed snow depth showed a negative bias.