In　order　to　better　understand　and　explore　the　soil　moisture　prediction　method,　this　paper　establishes　a　soil　moisture　prediction　model　based　on　soil　moisture　data　and　related　meteorological　data　from　2012-2021　in　Xilin　Gol　grassland　of　Inner　Mongolia,　while　keeping　the　current　grazing　strategy　unchanged,　based　on　Occam＇s　Razor　principle,　using　soil　moisture　and　existing　data　such　as　soil　evaporation　and　seasonal　changes.　Firstly,　the　ARIMA　(12,　1,　0)　model　was　developed　to　predict　the　precipitation　and　soil　evaporation　in　2022　and　2023.　Second,　according　to　the　transformation　relationship　between　evaporated　water　and　latent　heat　flux,　a　linear　model　of　soil　moisture　and　independent　variables　such　as　precipitation　and　evaporation　was　established,　and　the　expressions　of　different　soil　moisture　and　each　influencing　factor　were　obtained　by　fitting,　and　the　land　moisture　data　at　different　depths　were　calculated　as　the　prediction　results,　in　order　to　provide　technical　reference　for　further　exploration　of　soil　moisture　forecasting　and　better　meteorological　services　for　agriculture.　The　results　are　intended　to　provide　a　technical　reference　for　further　exploration　of　soil　moisture　forecasting　and　to　lay　the　foundation　for　better　meteorological　services　for　agriculture.　©　2023　SPIE.