[GRASS-user] [GRASSLIST:1178] how to use cloud cover with r.sun?

[Hamish]

Vishal Mehta wrote:

> But i noticed a major problem as i ran r.sun with the shadowing effect
> (-s) flag:
> Much of the centre of India ends up with 0 beam radiation!!
> See attached images- which are beam radiation output for the same day,
> with and without shadow. yellow color is zero beam radiation.
> 
> THE REASON FOR THIS (I think)
> I used the GTOPO30 (approx. 1km ORIGINAL res) elevation and derived
> slope and aspect as input to r.sun. HOWEVER, i never used the original
> GTOPO30 resolution- I just brought it into my current region, which is
> based on my base datasets of meterology from CRU at 0.5deg resoluton
> (approx. 50km? resolution), using r.in.bin  directly into my coarse
> region (0.5 deg). I think as a result, the shadowing effect in r.sun,
> which takes the max altitude from 4 nearest cells, is reaching across
> 50km(!) Hence with the Himalayas in the north and Western Ghats
> mountains in the west, most of central India gets 0 beam radiation!


I think you are correct, elevation in meters is probably being treaded
as degrees, so you get an unwanted 1852*60 vertical exaggeration.
Sounds plausible at least.

Solutions:

a) reproject your DEM into a meter-based projection (I couldn't tell you
which one), then reproject the results back to lat/lon. 

b) rescale your DEM into "degrees". e.g.:

r.mapcalc elevmap_scaled="elevmap / 1852*60"
  [maybe figure in cos(mean_lat)]

That will probably mess up the linke turbidity functions which I believe
takes physical elevation into account. (but I could be wrong)


Hamish