Thanks everyone,<br>
a few thoughts-<br>
1. Neither the r.sun manual, nor the r.slope.aspect manual, states that
the x,y and elevation units should be the same. In fact both manuals
say the elevation should be in meters! r.sun talks about how if the
maps are projected then it does'nt matter- but it does not talk of use
with lat-long system(nor does r.slope.aspect).<br>
<br>
do these manuals then just assume that we are using a projection where the x,y are in meters??<br>
<br>
2. Hamish and Markus are right - my study area covers several utm zones. <br>
<br>
3. At this point i'll try the dem, slope, aspect and r.sun in Lambert conformal conic projection with everest datum;<br>
or if that does not work, i'll try Hamish' idea of rescaling the elevation to degrees.<br>
<br>
I still cant get over the idea that GRASS cant handle correctly slope
generation in lat-long coordinates with elevation in meters!<br>
<br>
thanks again,<br>
vishal<br>
<br>
<br>
<br><div><span class="gmail_quote">On 8/25/06, <b class="gmail_sendername">Hamish</b> <<a href="mailto:hamish_nospam@yahoo.com" target="_blank" onclick="return top.js.OpenExtLink(window,event,this)">hamish_nospam@yahoo.com
</a>> wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">> But is it really possible that the r.sun module (i'm using Grass
6.0.1<br>> on Ubuntu Breezy) would treat elevation in degrees??<br><br>Yes, it is possible. Maybe even likely.<br><br>> And can you please tell me where the 1852*60 comes from?<br><br>There are 1852 meters in a nautical mile. There are 60 nautical miles in
<br>a degree of latitude. elev_in_meters / (1850.0 * 60) = elevation in deg<br><br>The width of a deg longitude is equal to width_lat*cos(lat).<br><br>If rescaling the elevation into degress be sure to /(1852.0*60). This
<br>ensures that the result will be a floating point value (as opposed to<br> /(1852*60) ).<br><br><br>> I guess i would have to creat a new location in UTM coordinates and<br>> then do the projecting back and forth you suggest..
<br><br>I think this is the saner solution, but I expect that a single UTM zone<br>is not an appropriate projection for working with an area the size of<br>India.<br><br>Markus wrote:<br>> This should be a projection covering the entire subcontinent
<br>> in a reasonable way - what are the official projection(s) for<br>> that area?<br><br>If you have problems finding one, you might ask for help on the PROJ.4<br>mailing list.<br><br><br><br>Hamish<br></blockquote>
</div><br>