I am getting very large negative values at the L outputs. Can it be related with high evotranspiration values? I’ve checked on the forum, somebody had a similar doubt, the answer point to check if the negative values were significant, and in my case, negative values can reach -1632.41 value on L output and no significan value on L_sum (you can see it on the image attached)
I’m not sure about the L values, but the grid pattern may be caused by using a “nearest neighbor” interpolation method on your DEM when you reprojected/resampled/etc to prepare it. “Bilinear”, “Cubic” and probably other interpolation methods do not produce this grid.
The very negative values in that upper section of the raster might be caused by the model not correctly handling nodata values from precipitation, quickflow or AET. Could you check those model outputs and see if any of them have nodata values where we’re seeing negative pixel values in L?