| Message/Author |
|
|
|
|
| List of features notes and updates to follow |
|
|
|
|
5.5q now stable version - Changes/notes include vectors/sweep widths/pitfilling algorithm.
Also new way of inputting hyrological data - means that we dont have to calibrate. Can run directly from a hydrograph. Divide hydrograph by number of grid cells and input directly (within the code) as the j_mean.
Been trying using different sweep widths
Comparison to surface flow vectors.
Try including WS smoothing, in low row example, this seemed to smooth the water surface really well, and make the velocities much better - need to re-visit
Seems to work well, but there does seem to be a problem with the smoothing up of too low cells.. in particular where sediment is added at the top - trying it at the moment with only peaks being decreased...
Managed to get a really nice compromise on the smoothing - by only allowing water depths to go up or down by 20% - makes things work really nice
Now found that problem with flow spliting and depositing too much on the intro cell was due to SS not being re-incorporated as suspended in the re-circulate mode. Need to also check for tracer input points.
ALso - found that there was 0 shear stress at input point leading to small mountain being developed at the input point (with vel tau calc.). This was due to l-r scan giving left out big total then re-input discharge in erodedepo then r-l giving right out a big total - so cancelling each other out at the input point! Simply added a line in erodedepo where the water is input in reach mode, to add the discharge to leftout.
There seems to be one problem using ws smoothing in that it can leave slightly - feathered edges - only in some scenarios - e.g. the feshie example. Now found that if you link the smoothing radius to the sweep width that gets rid of the rough edges somewhat.
5.7r stable and working well,
5.7s startng to see if down channels smoothing for lateral gradient
5.7s added soil erosion box, so can have creep and wash processes. explained on a new slope processes tab. Also move grass growth rate to veg tab.
5.7s Now have smooth downstream for lateral gradient. In present version, it smooths for 20 sweeps all directions, 20 downstream only, then the rest (up to 100 here). Seems to work well, though I am very concious of the need to fine tune this to different resolutions.
5.7t having a play with integrating SIBERIA.. may be some problems with the drainage area calculation part of the code - seems to have a significant impact upon the siberia operation.
That seems to have been tidied up by making a few subtle changes to the get area routine
5.7t discovered bug in max/mean slope calculating alogithms - had DX and root the wrong way around in a couple of places. Corrected, should have minimal effect on results |
|
|
| Back to top |
|
