The　economic　losses　caused　by　natural　disasters　to　human　society　are　increasing　year　by　year.　This　article　collects　the　loss　data　and　various　factors　caused　by　natural　disasters　in　China＇s　prefecture-level　cities　from　2014　to　2019,　and　uses　ArcGIS　software　to　visualize　them.　Then　use　analysis　of　variance　(ANOVA)　to　analyze　the　impact　of　various　factors　on　the　risk　of　natural　disasters　from　2014　to　2019.　Research　shows　that:　The　direct　economic　losses　caused　by　natural　disasters　and　the　affected　population　affected　by　natural　disasters　have　a　certain　degree　of　aggregation.　Disaster　type,　season　and　area　have　significant　influence　on　direct　economic　loss　and　affected　population　in　single-factor　analysis　of　variance.　The　main　effects　of　disaster　type　and　area　on　direct　economic　loss　and　disaster-affected　population　in　the　multivariate　analysis　of　variance　are　significant,　and　the　interaction　effect　is　also　significant.　©　2021　IEEE.