r.sun (mode 2): sawtooth bug in diffuse irradiation map
|Reported by:||hamish||Owned by:|
in investigating #2014 I noticed there's a bug in the calculation for diffuse sunlight when the sun is at low angles to the horizon. the overall trend seems ok, but there's a sawtooth pattern which follows the curvature of the lines of latitude.
The attached screenshots are for Ross Island in Antarctica, epsg:3286 bounds:
north: 1437000 south: 1341000 west: 17000 east: 119000 res: 100
DAY=060 r.sun -s elevin=etopo1.ross_island step=0.05 \ beam_rad=erebus.direct.$DAY diff=erebus.diffuse.$DAY day=$DAY