Glynn: > > The "curvature" table has a similar structure, but the colours aren't > > quite the same. > > > > The curvature table is: > > > > 0% black > > -0.01 blue > > -0.001 aqua > > -0.0001 cyan > > 0 white > > 0.0001 yellow > > 0.001 orange > > 0.01 red > > 100% magenta > > > > The rst/interp_float library generates the curvature > colour table > > programmatically, but the result is equivalent to: > > > > 0% purple > > -0.01 blue > > -0.001 indigo > > -0.00001 cyan > > 0 200:255:200 > > 0.00001 yellow > > 0.001 orange > > 0.01 red > > 100% 255:0:200 > > > > [The 100% case is close to magenta = 255:0:255]
Markus: > New table "rstcurv" backported, now available in 6.4, 6.5 > and 7. test: #spearfish r.slope.aspect elev=elevation.dem tcurv=elev.tcurv r.colors elev.tcurv color=curvature # r.colors elev.tcurv color=rstcurv observations: 1) "curvature" with its green midtones is more expressive than the more saturated rstcurv. A matter of preference I suppose. 2) with "d.vect quads type=boundary" you can really see the different datasets, esp. in the northern half. (elevation.dem is known to have other artifacts, e.g. grid visible in the slope map) 3) doing the same for elevation.10m works really well by comparison, you can really see the reduced curvature in the flatlands. Hamish _______________________________________________ grass-user mailing list grass-user@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/grass-user