Based on coKriging interpolation, the normalized vegetation index (NDVI), precipitation and river network density were used as covariates, and the groundwater depth of the Wengniute Banner in four phases of 2005—2017 was taken as the main variable. Then the groundwater depth in the study area was calculated. Using the improved hydrological response unit mode, the interpolated groundwater depth was spatially converted into vector data. And the traditional simple interpolation result analysis method was improved by using coKriging and minimum response unit to make it more suitable for actual groundwater convergence, and 551 minimum response units were obtained. Based on this, spatial and local autocorrelation analysis was performed. It can be observed that the depth of groundwater was obviously different spatially. The groundwater level was high in the west and low in the east, which was gradually affected by precipitation and rivers. The average depth of groundwater in time series did not change much, but it gradually showed a tendency of aggregation. 