[Redwulf__32]

I tried to write a similiar program, but yours work! Would have saved weeks and a lot of grey hairs in my Alpen map. - Seems that you set the water to the same height till a special height difference is reached. Would it be possible to lower this difference, best to "1"? I think you change the height of the riverbanks too. If the height jumps were lower, the "canyon-effect" might be lesser and some slope river sections might be corrected.

Here's how it works:
The map_c is inspected in 4X4 squares (200x200m). If a square is "all land" (all 16 pixel values > 127) or "all water" (all 16 pixel values =< 127) nothing is done. If a square is neither "all land" or "all water" (i.e. a coast) the 4 surrounding map_h pixels are set to the lowest value of the 4 (assuming the lowest value represents the water surface). This way inland water level is not pushed to 0m. There is no "fixed height difference", only values in the map_h file are used.

Hint: Do not use "Lake check" in MicroDem when building your map_h file. Doing so will push inland water level to 0m. Use "Lake check" on your map_c file though to get inland water.

FWIW: Made a correction table for MicroDem elevation according to info in this thread viewtopic.php?t=1714&postdays=0&postorder=asc&highlight=microdem+correction+table&start=45 See post by LSA on page 4.

Get it here: http://redwulf.adelby.dk/CoastAlign/Cor ... _map_h.dbf

The value 127 for distinguishing water/land is chosen since quite nice shallow water can be made by "fiddeling" with values just below 128 (land) in the map_c file.