Structure of GIS directories

The upper directory contains stability.aml, stuff2.aml, unstord.reclass, cohord.reclass, state2utm.prj, and directories for each of the 14 watersheds. It also contains some of the miscellany that accumulates anywhere.

Each watershed workspace contains coverages named LNDSLD (except for Calapooya) and BOUNDARY. If it has flood polygons (the Calapooya and the McKenzies), it has a workspace named FLOOD. If it has multiple resolution DEMs, there are workspaces named SCALE100FT and SCALE30FT. Otherwise, the DEMs remain in the main workspace.

Polygonal processing begins (more or less) with stuff.aml. Each watershed has a customizd version. You see that it creates three new items in LNDSLD.PAT, and that it tallies three categories:

even though the last two categories are aggregated in the tables. The flood coverage was processed as a unit before being split into three pieces.

Raster processing begins with the aml stability.aml. It is run from the workspace with the grids. The README file indicates the arguments used. It converts the initial DEM (always named ELV) to an image (raw.bil, raw.hdr, etc.) A script calls a fortran program to fill sinks, and a C program adds small slopes to flat spots and calculate contributing area / base. The result is read back in as the grids SLOPED and FLUX. SLOPED is elevation in millimeters, and flux is area / base in grid units. Intermediate files SLOPEDEG, TANSLOPE, SINSLOPE, COSSQ, AOBS, and WET1 are created before the actual models are run. The results are qust2, qust8, qust15 (Q-for-instability as phi = 33 and at various cohesion values, qust45 (at phi = 45 and cohesion = 0), and crticoh100 (critical cohesion at Q = 100). Intermediate grids QNT2, QNT8, QNT15, QNT45, and CNT100 can be deleted. The grids are reclassed to QUST2C, QUST8C, QUST15C, QUST45C, and CRITCOH100C, with values from 1 to 7. These are the only grids that are used in subsequent processing.

Up to this point landslide polygons and elevation grids have been processed completely independently. The aml stuff2.aml is called from the polygon directory to complete the processing. This is where things can be unclear on those watersheds that have two (standard and flood) polygon coverages. The processing is done for one of the coverages, then the files must be deleted before processing the second coverage.