Root restricting layer depth

I met a problem in processing root restricting layer depth. I’m going to substitute soil depth for root restricting layer depth. I downloaded data from some sample sites that recorded soil depth. I would like to know if it’s right to interpolate the soil depth to get the soil depth map. I checked the sample data and found that the sample data seems to be dissolved by the same value, not interpolated. How can I get the same data as the sample data.

Hi @frank -

It’s fine to use soil depth, and that layer can be created in different ways. The sample data is from a SOTER soil database, which provides a shapefile of polygons, with associated soil data that must be processed for each polygon. That’s why it looks like large areas with a single value - each polygon receives one value, and the data is coarse.

It sounds like you have point data for soil depth, and honestly I have not worked with that before so am not sure how to advise turning that into a soil depth raster. Interpolation is one way to do it, and we often use different interpolation methods for climate data, but it’s less clear to me how well that would represent differences in soil depth. I might also consider making something like Thiessen Polygons from the points, which might be an ok approximation, but in reality there’s a lot of geology defining where different depths begin and end.

Does anyone else have experience creating a soil depth raster from point data?

~ Stacie

Thanks @swolny . I think I need to review more references.

Another problem: There are some No Data pixels in the PAWC data of the area. Would it be ok to process No Data pixels as 0?

Hi @frank -

There are often NoData pixels in soil data where there’s water, glaciers etc. And yes, we should process them to some other value, otherwise the model results will have NoData values as well. I usually look at where they are and make a judgement call about them. Maybe they’re in places that are just solid rock, and a value of 0 might be fine. Otherwise, my default is to fill those holes in with the PAWC value that is most common around the NoData holes.

~ Stacie