Extractor Parameters Reference🔗

Each extractor has a setup_*_parameters() method that configures the extraction. This document lists the extractor-specific parameters for each.

Shared parameters (available on all extractors, not listed below): column_mapping, output_mapping, exclude_columns, output_columns, use_cache, partial_frequency

Check the output modes before you write a transform🔗

Several extractors take a parameter that changes the shape of what they return, not just its content — cropidExtractor's mode, FLMExtractor's and ZoningExtractor's postprocess, MRTSExtractor's mode, HistoricalScoreExtractor's detail_level, BaresoilExtractor's filter. They exist so a chain can get its input in the shape the next step wants, without glue.

So when you are about to write a transform, check the producing extractor's mode table first. The common mistake is to take the default shape and reshape it by hand: chaining CropID into Emergence looks like it needs a transform to collapse CropID's one-row-per-year output into a per-entity list of years — but mode: historical_season already emits exactly that, one row per entity.

Two things follow, and they remove most of the remaining glue:

Entity columns ride through. Every extractor stamps the input row's columns onto its results (normalize_with_metadata), which is why exclude_columns=["geometry"] exists at all. So id, geometry and crop survive a step, and a downstream depends_on receives them — no merge-back.
Reach for a transform when the grain genuinely changes, or when two independent branches have to be joined. Renaming a column is not that; use column_mapping (entity side) or output_mapping (result side).

⚠️ output_mapping is an extractor-level setting, applied via configure_output(). A workflow step cannot reach it — a step's setup block is a single method call, so it runs setup_*_parameters() or configure_output(), not both. In a workflow, rename on the entity side with settings.column_mapping, which is global to the run.

Foundational🔗

CoverageExtractor🔗

Method: setup_coverage_parameters() Module: earthdaily.agriculture.extractors.coverage_function

Parameter	Default	Description
`vegetation_index`	`"NDVI"`	Index type (NDVI, EVI, CVI, CVIN, GNDVI, LAI, NDWI, NDMI, S2REP)
`start_date`	`"2025-01-01"`	Coverage period start (YYYY-MM-DD)
`end_date`	`None`	Coverage period end (YYYY-MM-DD). `None` applies no upper bound — every image from `start_date` onward is returned, it is not capped at today
`clear_cover_min`	`90`	Minimum clear cover percentage
`clear_cover_max`	`100`	Maximum clear cover percentage
`use_specific_date`	`False`	Use exact date matching
`filter`	`"none"`	Filter mode: `none`, `duplicate`, `crop_coverage`
`delay`	`3`	Processing delay in days
`mask`	`"auto"`	Mask type: `auto`, `native`, `ACM`, `ML`
`recalibration`	`False`	Enable sensor recalibration
`historical_seasons`	`None`	List of prior years — required when `filter="crop_coverage"` (e.g. `[2024, 2023, 2022]`); the per-entity window is the entity's `start_date`/`end_date` month-day applied to each of those years

FLMExtractor🔗

Method: setup_flm_parameters() Module: earthdaily.agriculture.extractors.FLM_functions

Parameter	Default	Description
`vegetation_index`	`"NDVI"`	Index type for field-level maps
`map_format`	`None`	Output map format
`output_epsg`	`4326`	Output coordinate system EPSG code
`postprocess`	`"stats"`	Post-processing mode: `stats`, `links`, `file`, `histogram`
`extract_stats`	`False`	Extract statistics from maps
`directLinks`	`False`	Return direct download links
`clipping`	`"FieldBorder"`	Clipping method
`buffer`	`0`	Buffer around geometry (meters)
`skip_existing`	`True`	Skip already-downloaded maps
`output_path`	`None`	Override output directory

VegationTsExtractor🔗

Method: setup_vegetation_ts_parameters() Module: earthdaily.agriculture.extractors.VTS_functions

Parameter	Default	Description
`start_date`	`None`	Time series start (YYYY-MM-DD)
`end_date`	`None`	Time series end (YYYY-MM-DD)
`vegetation_index`	`"NDVI"`	Index type
`is_extrapolated`	`True`	Enable end-of-curve extrapolation
`limit`	`3000`	Max data points per request
`historical_years`	`10`	Number of historical years to compare
`extraction_mode`	`"period"`	Extraction mode — see below
`target_dates`	`None`	List of specific dates to extract
`kpi_filter`	`None`	KPI aggregation filter

extraction_mode values:

Mode	Description
`period`	Continuous extraction between `start_date` and `end_date`. Default.
`specific_dates`	Extract only on the dates listed in `target_dates`. Requires `target_dates`.
`windows`	Per-entity windowed extraction; `start_date` / `end_date` (or per-entity overrides) define the window.

MRTSExtractor🔗

Method: setup_mrts_parameters() Module: earthdaily.agriculture.extractors.VTS_functions

Parameter	Default	Description
`start_date`	`"2025-05-01"`	Time series start (YYYY-MM-DD)
`end_date`	`"2025-10-15"`	Time series end (YYYY-MM-DD)
`sensors`	`None`	Sensor filter (Sentinel-2, Landsat, etc.)
`vegetation_index`	`"NDVI"`	Index type
`aggregation`	`"average"`	Spatial aggregation method
`smoothing_method`	`"Whittaker"`	Smoothing algorithm
`apply_denoiser`	`True`	Enable denoising
`apply_end_of_curve`	`True`	Enable extrapolation
`clear_cover_min`	`100`	Minimum clear cover percentage
`output_saturation`	`True`	Output saturation flag
`extract_raw_datasets`	`True`	Include raw (unsmoothed) data
`compute_temporal_consistency`	`True`	Run temporal consistency check
`temporal_consistency_threshold`	`None`	Custom consistency threshold
`mode`	`"full"`	Output mode — see below
`historical_years`	`10`	Number of historical years
`kpi_filter`	`None`	KPI aggregation filter

mode values:

Mode	Description
`full`	Returns raw + smoothed values, with denoiser, end-of-curve extrapolation, and temporal-consistency flags. Default.
`raw`	Returns only raw per-image values (skips smoothing / denoising stages).

WeatherExtractor🔗

Method: setup_weather_parameters() Module: earthdaily.agriculture.extractors.weather_functions

Parameter	Default	Description
`weather_type`	`"HISTORICAL_DAILY"`	Weather data type
`weather_parameters`	`"none"`	Specific weather parameters to extract
`kpi_filter`	`None`	KPI aggregation filter

cropidExtractor🔗

Method: setup_cropid_parameters() Module: earthdaily.agriculture.extractors.cropid_functions

Parameter	Default	Description
`begin_year`	`2020`	First year of crop history
`end_year`	`2025`	Last year of crop history
`mask_type`	`"EndSeason"`	Mask type: `EndSeason`, `InSeason`, or `PreSeason`
`limit_nb_crop`	`1`	Number of top crops to return
`crop_mask_percent`	`50`	Minimum crop mask percentage
`mode`	`"history"`	Extraction mode — see below

mode values:

Mode	Shape	Description
`history`	long-form	One row per (year, crop) across the window. Default.
`year`	long-form	Filtered to the entity's `crop`; rows for matching years only.
`historical_season`	one row, one string column	Comma-separated string of matching years (e.g. `"2020,2022,2024"`).
`full_history`	wide-form	One row per entity, one column per year in `begin_year..end_year` (NaN if no data). Multi-crop years collapse to the top crop by `cropMaskPercent`. Convenient as an entity-level summary to merge alongside other extractors.

ZoningExtractor🔗

Method: setup_zoning_parameters() Module: earthdaily.agriculture.extractors.zoning_functions

Parameter	Default	Description
`num_zones`	`5`	Number of management zones
`output_epsg`	`4326`	Output coordinate system EPSG code
`postprocess`	`"stats"`	Post-processing mode: `stats`, `stats_geo`, `links`, `file`
`map_format`	`None`	Output map format
`output_path`	`None`	Override output directory
`skip_existing`	`True`	Skip already-processed entities
`directLinks`	`False`	Return direct download links

postprocess values:

Mode	Description
`stats`	One field-level row per entity: variability, productivity/variability indices, and `zone_{n}_*` columns. Default.
`stats_geo`	One row per zone, carrying that zone's geometry (segments merged into a `GEOMETRYCOLLECTION`) plus its mean/max/min/area, with the field-level stats repeated on every row.
`links`	Direct download links (requires `directLinks=True`, set automatically).
`file`	Downloads the map; requires `map_format` and `output_path`.

LocationBasedBorderExtractor🔗

Method: setup_location_based_border_parameters() Module: earthdaily.agriculture.extractors.location_based_border_functions

Wraps the Geosys /field-borders/v1/AutomaticBoundary endpoint. Consumes a DataFrame with a Point WKT in the geometry column and returns the field polygon (WKT) containing that point.

Parameter	Default	Description
`simplified_geom`	`True`	If `True`, the API returns a simplified field geometry (lower shape-point count). Maps to the `simplified_geom` query parameter.

Required entity fields: id, geometry (must be a Point WKT — polygons are rejected with a clear error so callers can pre-process to a centroid via core.geometry.get_centroid_wkt()).

Output columns: entity_id, point_geometry (input echo), polygon_geometry (returned field border WKT), plus any flat scalar properties returned by the API (e.g. area_ha, sourceId).

Regional🔗

RegionalExtractor🔗

Method: setup_regional_parameters() Module: earthdaily.agriculture.extractors.regional_ts_extractor

Parameter	Default	Description
`index`	`"vegetation-vigor-index"`	Regional index type
`start_date`	`"2025-01-01"`	Period start (YYYY-MM-DD)
`end_date`	`None`	Period end (defaults to Dec 31 of current year)
`fillyeargap`	`False`	Fill year gaps in data
`idblock`	`None`	Block identifier
`idpixeltype`	`None`	Pixel type identifier
`indicatorTypeIds`	`None`	Indicator type IDs (auto-set to [1] for VVI)

Valid index values: vegetation-vigor-index, daily-precipitation, soil-moisture, min-temperature, max-temperature, average-temperature, surface-temperature

Valid indicatorTypeIds: 1 (VVI), 2 (Weather ECMWF), 3 (Weather AROME France), 4 (Forecast ECMWF), 5 (Forecast GFS)

10 (Weather Reanalysis) and 11 (Rainfall Estimates HI-RES / CHIRPS) appear in the API documentation but are not accepted by this extractor — passing them raises ValueError. Widen valid_indicatorTypeIds in regional_ts_extractor.py first.

Crop Development & Stressors🔗

DiseaseExtractor🔗

Method: setup_disease_parameters() Module: earthdaily.agriculture.processors.processor_disease_risk_functions

Parameter	Default	Description
`start_date`	`None`	Risk assessment period start
`end_date`	`None`	Risk assessment period end

EmergenceExtractor🔗

Method: setup_emergence_parameters() Module: earthdaily.agriculture.processors.processor_emergence_functions

Parameter	Default	Description
`emergence_type`	`"INSEASON"`	Detection type — see below
`season_duration`	`120`	Season length in days
`season_start_day`	`1`	Season start day of month
`season_start_month`	`4`	Season start month (1-12)
`year`	`2025`	Target year
`data_source`	`"LR"`	Data source: `LR` or `MR`
`publish_af`	`False`	Publish analytic feature

emergence_type values:

Type	Description
`INSEASON`	Single emergence date for the current season (`emergence_date`, `emergence_status`, `confirmation_status`). Default.
`HISTORICAL`	Emergence dates for the last 5 years (`emergence_year_1`..`emergence_year_5`) plus `historical_average_emergence`. Also accepts an optional per-entity `historical_seasons` column — see below.
`DELAY`	Current-season emergence date compared to the historical average (`emergence_date`, `average_emergence_date`, `emergence_delay`).

historical_seasons — a per-entity input, not a setup parameter. In HISTORICAL mode the extractor reads a historical_seasons value off each entity row (via column_mapping, like any other entity field): the calendar years that field actually grew the crop, as "2020,2022,2024" or [2020, 2022, 2024]. Years outside the list are nulled out of emergence_year_N, and an extra avg_emergence_matching_years column (MM-DD) holds the average recomputed over the years kept. historical_average_emergence — the raw API average over all five seasons — is left untouched either way.

It fails silently. The column is optional, so when it is absent (or mapped under the wrong name) nothing is filtered and nothing is raised: all five years come back and the output looks correct. If you are chaining CropID into Emergence to get this, check that the names line up — cropidExtractor's historical_season mode emits historical_season (singular) while this extractor reads historical_seasons (plural), so the chain needs a column_mapping entry.

HarvestExtractor has the same mechanism in HISTORICAL_HARVEST mode, producing avg_harvest_matching_years.

sowing_date is not read in any mode — the season window comes from season_start_month / season_start_day / season_duration / year here.

GreennessExtractor🔗

Method: setup_greenness_parameters() Module: earthdaily.agriculture.processors.processor_greenness_functions

Parameter	Default	Description
`season_duration`	`120`	Season length in days
`season_start_day`	`1`	Season start day of month
`season_start_month`	`4`	Season start month (1-12)
`year`	`2025`	Target year
`sowing_date`	`"2025-04-01"`	Default sowing date (YYYY-MM-DD)
`data_source`	`"LR"`	Data source: `LR` or `MR`
`publish_af`	`False`	Publish analytic feature

HarvestExtractor🔗

Method: setup_harvest_parameters() Module: earthdaily.agriculture.processors.processor_harvest_functions

Parameter	Default	Description
`harvest_type`	`"INSEASON_HARVEST"`	Detection type — see below
`season_duration`	`120`	Season length in days
`season_start_day`	`1`	Season start day of month
`season_start_month`	`4`	Season start month (1-12)
`year`	`2025`	Target year
`data_source`	`"LR"`	Data source: `LR` or `MR`
`publish_af`	`False`	Publish analytic feature

harvest_type values:

Type	Description
`INSEASON_HARVEST`	Single harvest date + status for the current season (`harvest_date`, `harvest_status`). Default.
`HISTORICAL_HARVEST`	Harvest dates for the last 5 years (`harvest_year_1`..`harvest_year_5`) plus `historical_harvest_average`.
`HARVEST_READINESS`	Estimated harvest-readiness date and an `is_ready` boolean.

PlantedExtractor🔗

Method: setup_planted_parameters() Module: earthdaily.agriculture.processors.processor_plantedarea_functions

Parameter	Default	Description
`processor_mode`	`"PLANTED_AREA"`	Processing mode — see below
`emergence_date`	`None`	Known emergence date override
`threshold`	`120`	Detection threshold
`control_threshold`	`4`	Control threshold
`publish_af`	`False`	Publish analytic feature

processor_mode values:

Mode	Description
`PLANTED_AREA`	Returns `planted_area_m2` and `planted_percentage` for the field. Default.
`CONTROL`	Returns control-check output: `difference`, `control_threshold`, `control_result` (used to validate planting status against a known baseline).

InSeasonMonitoringExtractor🔗

Method: setup_inseason_monitoring_parameters() Module: earthdaily.agriculture.processors.processor_inseason_monitoring_functions

Parameter	Default	Description
`season_duration`	`120`	Season length in days
`season_start_day`	`1`	Season start day of month
`season_start_month`	`4`	Season start month (1-12)
`year`	`"2025"`	Target year (string)
`data_source`	`"LR"`	Data source: `LR` or `MR` (one per run)

ChangeIndexExtractor🔗

Method: setup_change_index_parameters() Module: earthdaily.agriculture.processors.processor_change_index_functions

Compares a reference image (one per entity, via reference_date) against the nearest prior image within a configurable look-back window to flag significant change on a field. Useful for harvest detection, stress events, tillage, and other rapid changes.

Parameter	Default	Description
`map_type`	`"NDVI"`	Vegetation index — one of `NDVI`, `EVI`, `CVI`, `GNDVI`, `NDWI`
`collections`	`None`	Satellite collections list (default `["Sentinel-2"]` when `None`)
`max_period_reference`	`7`	Max days back from `reference_date` for the reference image
`max_period_previous`	`15`	Max days before the reference image for the previous image
`min_period_previous`	`5`	Min days before the reference image for the previous image
`same_sensor`	`False`	If `True`, require the same sensor for both images
`parameter_profile`	`"change_index_v1"`	API parameter profile name
`publish_af`	`False`	If `True`, include `field_id` in the request so the result is registered on the platform. Set this to `True` — see the note below.

publish_af=False currently fails on this extractor. The change-index API rejects a payload without field_id:
422 {"detail":[{"type":"missing","loc":["body","field_id"],"msg":"Field required"}]}
so the default cannot complete a call. Pass publish_af=True until that is fixed — note this registers the result on the platform, and the entity must carry an id.

This is specific to ChangeIndex; the other processors exposing publish_af are unaffected and their False default works. A fix is ticketed on the API side, because requiring a platform-managed field_id conflicts with this package's geometry-first design.

Required entity fields: id, geometry, reference_date (column-mapping aware). Optional entity fields: crop, sowing_date.

Output columns: entity_id, reference_date, status, plus any change metrics returned by the API for the selected map_type.

Risk Management🔗

HistoricalScoreExtractor🔗

Method: setup_historical_score_parameters() Module: earthdaily.agriculture.processors.processor_score_functions

Parameter	Default	Description
`season_duration`	`120`	Season length in days
`season_start_day`	`1`	Season start day of month
`season_start_month`	`4`	Season start month (1-12)
`threshold_start`	`0.7`	Score threshold start
`year`	`2025`	Target year
`historical_seasons`	`None`	List of years to compare
`data_source`	`"LR"`	Data source: `LR` or `MR`
`detail_level`	`"full"`	Output detail level — see below
`publish_af`	`False`	Publish analytic feature

detail_level values:

Level	Description
`full`	Summary metrics (`average_potential_score`, `olympic_mean_potential_score`, `standard_deviation`, `risk_score`) plus per-season `potential_score_<year>` and `season_break_<year>` columns. Default.
`summary`	Summary metrics only — one row per entity, no per-season detail.

InseasonScoreExtractor🔗

Method: setup_inseason_score_parameters() Module: earthdaily.agriculture.processors.processor_score_functions

Parameter	Default	Description
`season_duration`	`120`	Season length in days
`season_start_day`	`1`	Season start day of month
`season_start_month`	`4`	Season start month (1-12)
`nb_historical_year`	`1`	Number of historical years
`threshold_start`	`0.7`	Score threshold start
`historical_seasons`	`None`	List of years to compare
`data_source`	`"LR"`	Data source: `LR` or `MR`
`detail_level`	`"full"`	Output detail level — see below
`publish_af`	`False`	Publish analytic feature

detail_level values:

Level	Description
`full`	Summary metrics plus per-season `potential_score_<year>` and `season_break_<year>` columns. Default.
`summary`	Summary metrics only — one row per entity, no per-season detail.

ZARCExtractor🔗

Method: setup_zarc_parameters() Module: earthdaily.agriculture.processors.processor_zarc_functions

Parameter	Default	Description
`crop`	`"OTHERS"`	Crop type for ZARC lookup
`nb_days_sowing_emergence`	`20`	Days from sowing to emergence
`soil_type`	`None`	Soil type classification
`cycle`	`None`	Crop cycle type

Sustainability🔗

BaresoilExtractor🔗

Method: setup_baresoil_parameters() Module: earthdaily.agriculture.processors.processor_baresoil_function

Parameter	Default	Description
`season_duration`	`120`	Season length in days
`season_start_day`	`1`	Season start day of month
`season_start_month`	`4`	Season start month (1-12)
`year`	`2025`	Target year
`filter`	`"summary"`	Output mode: `summary` or `detailed`
`publish_af`	`False`	Publish analytic feature

Quick Reference — All Params by Extractor🔗

Legend: S = season window params (season_duration, season_start_day, season_start_month, year)

Extractor	Dates	Index / Type	Season (S)	Data Source	Other key params
CoverageExtractor	`start_date`, `end_date`	`vegetation_index`			`clear_cover_min`, `filter`, `mask`, `delay`, `historical_seasons`
FLMExtractor		`vegetation_index`			`clipping`, `buffer`, `output_epsg`, `postprocess`, `extract_stats`
VegationTsExtractor	`start_date`, `end_date`	`vegetation_index`			`extraction_mode`, `target_dates`, `historical_years`, `kpi_filter`
MRTSExtractor	`start_date`, `end_date`	`vegetation_index`		`sensors`	`smoothing_method`, `clear_cover_min`, `mode`, `historical_years`, `kpi_filter`
WeatherExtractor					`weather_type`, `weather_parameters`, `kpi_filter`
cropidExtractor					`begin_year`, `end_year`, `mask_type`, `limit_nb_crop`, `mode`
ZoningExtractor					`num_zones`, `output_epsg`, `postprocess`
RegionalExtractor	`start_date`, `end_date`	`index`			`idblock`, `idpixeltype`, `indicatorTypeIds`, `fillyeargap`
DiseaseExtractor	`start_date`, `end_date`
EmergenceExtractor			S	`data_source`	`emergence_type`
GreennessExtractor			S	`data_source`	`sowing_date`
HarvestExtractor			S	`data_source`	`harvest_type`
PlantedExtractor					`processor_mode`, `emergence_date`, `threshold`, `control_threshold`
InSeasonMonitoringExtractor			S	`data_source`
ChangeIndexExtractor	`reference_date` (per entity)	`map_type`		`collections`	`max_period_reference`, `max_period_previous`, `min_period_previous`, `same_sensor`, `parameter_profile`, `publish_af`
HistoricalScoreExtractor			S	`data_source`	`threshold_start`, `historical_seasons`, `detail_level`
InseasonScoreExtractor			S	`data_source`	`nb_historical_year`, `threshold_start`, `historical_seasons`, `detail_level`
ZARCExtractor					`crop`, `nb_days_sowing_emergence`, `soil_type`, `cycle`
BaresoilExtractor			S		`filter`

Common Param Patterns🔗

Season window params (shared by 7 extractors)🔗

season_duration=120,      # season length in days
season_start_day=1,       # day of month (1-31)
season_start_month=4,     # month (1-12, 4=April)
year=2025,                # target year

Used by: Emergence, Greenness, Harvest, InSeasonMonitoring, HistoricalScore, InseasonScore, Baresoil

Date range params🔗

start_date="2025-01-01",  # YYYY-MM-DD
end_date="2025-12-31",    # YYYY-MM-DD or None

Used by: Coverage, VegationTs, MRTS, Weather (via entity), Disease, Regional

KPI filter🔗

kpi_filter=None,           # dict with KPI aggregation rules

Used by: VegationTs, MRTS, Weather

See here for more details.

Data source🔗

data_source="LR",          # "LR" (low-res) or "MR" (medium-res)

A single source per run — the API query carries one dataSource value. Run the extractor twice if you need both resolutions.

Used by: Emergence, Greenness, Harvest, HistoricalScore, InseasonScore, InSeasonMonitoring