analysis-dag.md

Since the dataset with highest resolution is the DEM, we will align our datasets to it.

Notes

• In our DAG, rectangular nodes with squared edges represent data, whereas nodes with rounded edges represent computing tasks.
• By align, we denote a computing task consists of:
  • Cropping an input raster to match the extent of a reference raster, and
  • Aligning the input raster to match the reference of a reference raster
• Saving .nc files aligned to a finer resolution is verbose and space inefficient, so we will just crop them first, and align them in runtime within the STREAM step (thus avoiding to store the files)

graph TD
DEM[DEM tif] --> CROP(Crop to watershed extent);
WSHED[Watershed shp] --> CROP;
CROP --> DEMC[Cropped DEM];

DEMC --> ALIGNL(Align to Cropped DEM);
LULC[LULC tif] --> ALIGNL;
ALIGNL --> LULCA[Aligned LULC];

LULCA --> MAPC(Map to crop factor);
MAPC --> CROPF[CROPF];

DEMC --> CROPT(Crop to watershed extent);
TEMP[TEMP nc] --> CROPT;
CROPT --> TEMPA[Cropped TEMP];

DEMC --> CROPP(Crop to watershed extent);
PREC[PREC nc] --> CROPP;
CROPP --> PRECA[Cropped PREC];

DEMC --> FILL(Fill pits);
FILL --> DEMF[Filled DEM];

DEMC --> ALIGNW(Align to Cropped DEM);
WHC[WHC tif] --> ALIGNW;
ALIGNW --> WHCA[Aligned WHC];

DEMF --> STREAM[STREAM];
CROPF --> STREAM;
WHCA --> STREAM;
TEMPA --> STREAM;
PRECA --> STREAM;
STREAM --> FLOW[Flow];