Configuration Examples

Minimal (linear interpolation)

raster_path: "examples/sample_data/soil.tif"
climate_csv: "examples/sample_data/climate.csv"
output_dir: "outputs"
roi:
  type: bbox
  xmin: -120.5
  ymin: 34.0
  xmax: -118.0
  ymax: 35.5
model_params:
  v_min: 0.0
  v_max: 1.0
  t_min: 10.0
  t_max: 35.0
  r_min: 100.0
  r_max: 800.0
  w_v: 0.4
  w_t: 0.3
  w_r: 0.3
climate:
  strategy: spatial
  interpolation_method: linear   # default — fast, no extra deps
  fallback_to_mean: true
max_cells: 250

Kriging with uncertainty

Adds mean_temp_krig_std / total_rain_krig_std columns and (with uncertainty_samples > 0) score_ci_low / score_ci_high columns.

raster_path: "examples/sample_data/soil.tif"
climate_csv: "examples/sample_data/climate.csv"
output_dir: "outputs"
roi:
  type: bbox
  xmin: -120.5
  ymin: 34.0
  xmax: -118.0
  ymax: 35.5
model_params:
  v_min: 0.0
  v_max: 1.0
  t_min: 10.0
  t_max: 35.0
  r_min: 100.0
  r_max: 800.0
  w_v: 0.4
  w_t: 0.3
  w_r: 0.3
  uncertainty_samples: 500   # Monte Carlo draws for score CI
climate:
  strategy: spatial
  interpolation_method: kriging
  variogram_mode: standard
  fallback_to_mean: true
max_cells: 250

Set variogram_mode: extended when you want TerraFlow to try additional nested variogram candidates before selecting the best kriging model. Extended mode is more expensive than standard kriging because each nested candidate fits a custom variogram before LOOCV; prefer standard mode for large station networks unless nested structures are specifically needed.

IDW (fast, no uncertainty)

raster_path: "examples/sample_data/soil.tif"
climate_csv: "examples/sample_data/climate.csv"
output_dir: "outputs"
roi:
  type: bbox
  xmin: -120.5
  ymin: 34.0
  xmax: -118.0
  ymax: 35.5
model_params:
  v_min: 0.0
  v_max: 1.0
  t_min: 10.0
  t_max: 35.0
  r_min: 100.0
  r_max: 800.0
  w_v: 0.4
  w_t: 0.3
  w_r: 0.3
climate:
  strategy: spatial
  interpolation_method: idw
  fallback_to_mean: true
max_cells: 250

See Climate Configuration for full field reference.