Hi @XiaoYu -
It is fine to look at wyield.tif to assess the patterns of water yield across the landscape, we just don’t recommend using the individual pixel values.
I would say that the Annual Water Yield model is most appropriate for places that are dominated by rainfall. The model itself doesn’t include any of the dynamics of melting snow and ice, and because it operates at an annual average timescale, you won’t see any seasonality to the water flow that might be associated with the snow and ice melt. We do have a Seasonal Water Yield model that operates at a monthly time scale, but again it’s not really made for places dominated by snow. Here’s a forum post that references a paper that might be useful about snow and the SWY model (key takeaway: SWY doesn’t perform well in areas dominated by snow, and the authors used a separate model for snow melt).
That said, did you include snow and ice melt as part of the annual average precipitation raster you used as input? That would be one way to at least make sure that the source of additional water is included in the model.
~ Stacie