how can I integrate continuous data and static data using random forest machine learning model?

I am using random forest regression model to predict groundwater level changes. I am using continuous inputs (timeseries data) such as GRACE, Precipitation, Maximum temperature, Minimum temperature, NDVI as well as static data such as land elevation, hydraulic conductivity, slope, sand percent. When I added static inputs to continuous inputs, the model gave high importance to static inputs and neglected continuous inputs. How I can fix this problem.


I got a prediction but the problem in feature importance the model gave high importance to static inputs and no importance for continuous inputs which is wrong. how can I fix this?