NAME

KEYWORDS

SYNOPSIS

Flags

-d

Ignore input=, import the bundled Tonle Sap demonstration dataset for sector= instead

DESCRIPTION

v.tonlesap.model implements a multi-criteria pairwise-comparison outranking model to score the suitability of Tonle Sap (Cambodia) villages for one of three sectors: irrigation, aquaculture or potable water. It is a native GRASS vector port of the standalone TSAPMODEL scripts (TLI.py, TLA.py, TLP.py), written for IWMI.

The module reads three attribute columns (the outranking criteria) from the input vector points map, compares every pair of villages for each criterion, and accumulates a signed score per village weighted by weights and the better direction (more is better or less is better). The result is rescaled to 0-1 and written to a new score_column (default tsap_score) in output.

An optional screening step removes villages that do not satisfy one or more threshold conditions (e.g. TOPOZONE,le,3) from the output map — screening does not affect how the score itself is computed, it only filters which villages appear in the final result, matching the behaviour of the original scripts.

Criteria selection

Which attribute columns are used depends on sector:

• irrigation (all three required):
  • crit1: 0=MA_INDACT, 1=MA_PROXROAD
  • crit2: 0=TAGAP_DRY, 1=TAGAP_WET
  • crit3: 0=SW_PROXRIV, 1=SW_PONDS, 2=GW_DWELL, 3=GW_BOREW, 4=IRRI_SCH, 5=IRRI_HEAD
• aquaculture (crit1/crit2 required, crit3 ignored):
  • crit1: 0=MA_INDACT, 1=MA_PROXROAD
  • crit2: 0=SW_PROXRIV, 1=SW_PONDS
  • 3rd criterion is always FISHPRO (compulsory)
• potable (crit1/crit2/crit3 ignored, always uses all three): WA_AWAY, WS_UNSAFE, WT_UNTREAT

weights and better always take exactly 3 comma-separated values, in the same order as the resolved criteria above.

Demonstration dataset

The addon ships the original TSAPMODEL tablefile.csv for each sector (surveyed in Indian 1960 / UTM zone 48N, EPSG:3148). Pass the -d flag to import the bundled dataset for the chosen sector on the fly and use it as input instead of input= — this recreates the original demonstration mapping without needing any external data. The current GRASS project must be in EPSG:3148 (create one with grass -c EPSG:3148 <GISDBASE>/<project> -e if needed).

NOTES

The pairwise comparison is O(n^2) per criterion (n = number of input features), vectorised with NumPy. Very large village counts will still be slow; consider pre-filtering the input vector if performance becomes an issue.

EXAMPLES

Reproduce the original TLI.py self-test (irrigation, proxroad + wet + proxriv) using the bundled demonstration dataset:

grass -c EPSG:3148 ~/grassdata/tonlesap_demo -e
grass ~/grassdata/tonlesap_demo/PERMANENT --exec \
  v.tonlesap.model -d sector=irrigation output=irrigation_score \
  crit1=0 crit2=1 crit3=0 weights=1.0,1.0,1.0 better=m,l,l

Run the aquaculture model on an existing vector, screening out villages above elevation 5 and with TOPOZONE greater than 3:

v.tonlesap.model input=villages output=aqua_score sector=aquaculture \
  crit1=0 crit2=1 weights=1.0,1.0,1.0 better=m,l,m \
  screening=ELEVATION,ge,5,TOPOZONE,le,3

Run the potable-water model (fixed 3-column criteria set, no crit1/2/3):

v.tonlesap.model input=villages output=potable_score sector=potable \
  weights=1.0,1.0,1.0 better=m,m,l

SEE ALSO

AUTHORS

Original TSAPMODEL: github.com/YannChemin/TSAPMODEL