🌲 `lands`¶

U.S. federal public lands data covering USDA Forest Service, National Park Service, and Bureau of Land Management units. Includes land management boundaries, timber sale program contracts, forest inventory estimates, NPS visitation statistics, and Office of Natural Resources Revenue (ONRR) mineral royalty payments. Enables analysis of extractive industry activity on federal land, tourism economics, and timber/paper company revenue risk via SEC filings.

30 datasets · 214 columns

national_forests · table¶

USDA Forest Service National Forest and Grassland unit boundaries. ~175 units across 10 USFS regions. Reference/dimension table — FK target for timber_sales.forest_code via forest_id. Updated annually post-fire-season (October). geometry_wkt carries the unit's own WGS84 boundary polygon, sourced from the same ArcGIS query at ingest time (no separate spatial join needed to plot or intersect an individual forest); for federal-fee ownership share by county across all agencies, use lands.padus_federal_fee_lands / lands.county_federal_fee_land_share instead, since PAD-US's own_type/mang_name lets that dataset be filtered or compared across USFS, BLM, and NPS in one place.

Column	Type	Null	Description
`forest_id`	string	no	USDA FS unit ID (e.g., '0501')
`forest_name`	string	yes	Unit name (e.g., 'Angeles National Forest')
`region`	string	yes	USDA FS region number (01–10)
`gross_acres`	double	yes	Total unit area in acres (GIS measurement)
`geometry_wkt`	string	yes	Simplified boundary WKT polygon (for display only)

timber_sales · table¶

USDA Forest Service timber harvest activities from FACTS EDW. One row per completed harvest activity. Joins to national_forests via forest_code and to geo.states via state_abbr. Units_accomplished and uom describe harvest volume (CCF, MBF, ACRE, etc.). Enables analysis of federal timber activity by forest, district, and year. Also carries cost_per_uom (USD per unit of measure) for harvest cost analysis. Source: EDW_TimberHarvest_01/MapServer/{layer} where layer is dispatched per year by UsfsTimberLayerResolver — layer 11 covers 2021+, layer 0 covers 2011-2020, and layers 1-7 could in principle reach as far back as 1820, but this table is only ever fetched from GOVDATA_START_YEAR forward (see observedCoverage below for the real floor actually loaded) — the 1820 figure describes what the source's older layers contain, not what's in this table. The service hard-caps maxRecordCount at 2000 regardless of the requested resultRecordCount (a single year can hold 4x that); UsfsFactsTransformer pages on exceededTransferLimit via resultOffset rather than making one request per year. All layers publish the identical 71-field schema.

Column	Type	Null	Description
`facts_id`	string	no	FACTS activity unique ID (PK)
`sale_name`	string	yes	Timber sale name
`fy_completed`	integer	no	Fiscal year completed (partition key)
`fy_awarded`	integer	yes	Fiscal year awarded
`forest_code`	string	yes	USFS administrative forest code (FK to national_forests.forest_id)
`forest_name`	string	yes	Administrative forest name
`district_code`	string	yes	Ranger district code
`district_name`	string	yes	Ranger district name
`state_abbr`	string	yes	2-character state abbreviation
`gis_acres`	double	yes	Activity area in acres (GIS measurement)
`activity_code`	string	yes	FACTS activity code
`activity_name`	string	yes	FACTS activity name
`treatment_type`	string	yes	Treatment category (Salvage / Regeneration / etc.)
`units_accomplished`	double	yes	Harvest volume or area units accomplished
`uom`	string	yes	Unit of measure (CCF / MBF / ACRE / etc.)
`cost_per_uom`	double	yes	Cost per unit of measure (USD)

forest_inventory · table¶

USDA Forest Service Forest Inventory and Analysis (FIA) state-level estimates derived from per-state {ST}_COND.csv files inside the FIA datamart's per-state ZIP archives (apps.fs.usda.gov/fia/datamart/CSV/{state}_CSV.zip). One row per inventory_year × state × forest_type_group × ownership_class. Filtered to COND_STATUS_CD=1 (accessible forest land). Basal area is a CONDPROP_UNADJ-weighted average of BALIVE. Joins to geo.states via state_fips. ETL fans out across states; cached per-state archives in $GOVDATA_CACHE_DIR/lands/ are shared across workers. Materialization keeps state as a row column, not a partition column, so that Iceberg per-partition compaction merges the 55 per-state ETL flushes into a single nationwide parquet.

Column	Type	Null	Description
`inventory_year`	integer	no	FIA inventory year (INVYR)
`state_fips`	string	no	2-digit FIPS code identifying the primary state (e.g., '06' for California)
`forest_type_group`	string	yes	Forest type group name (e.g., 'Oak / hickory group', 'Douglas-fir group')
`ownership_class`	string	yes	FIA OWNGRPCD group name: National Forest (10), Other Federal (20), State and local government (30), Private (40). Those four are the whole OWNGRPCD domain — state, local and other non-federal public are combined under 30, and the finer split lives in OWNCD, which this table does not carry.
`basal_area_sqft`	double	yes	CONDPROP_UNADJ-weighted average of BALIVE (sq ft per acre)

forest_metrics · table¶

USDA FIA tree-level metrics derived from per-state {ST}_TREE.csv joined to {ST}_COND.csv (both inside the per-state ZIP archive at apps.fs.usda.gov/fia/datamart/CSV/{state}_CSV.zip). One row per inventory_year × state × forest_type_group × ownership_class. Columns are CONDPROP_UNADJ-weighted averages across live trees (STATUSCD=1) in accessible forest conditions (COND_STATUS_CD=1). Join to forest_inventory on (state_fips, inventory_year, forest_type_group, ownership_class) for combined basal-area + volume + carbon analysis. ETL fans out across states.

Column	Type	Null	Description
`inventory_year`	integer	no	FIA inventory year (INVYR)
`state_fips`	string	no	2-digit FIPS code identifying the primary state (e.g., '06' for California)
`forest_type_group`	string	yes	Forest type group name (e.g., 'Oak / hickory group', 'Douglas-fir group')
`ownership_class`	string	yes	FIA OWNGRPCD group name: National Forest (10), Other Federal (20), State and local government (30), Private (40). Those four are the whole OWNGRPCD domain — state, local and other non-federal public are combined under 30, and the finer split lives in OWNCD, which this table does not carry.
`trees_per_acre`	double	yes	CONDPROP-weighted average of TPA_UNADJ across live trees (trees per acre)
`live_volume_cuft`	double	yes	CONDPROP-weighted average live cubic-foot volume per acre (VOLCFNET × TPA_UNADJ)
`carbon_stock_tons`	double	yes	CONDPROP-weighted average above-ground carbon stock, tons per acre (CARBON_AG × TPA_UNADJ / 2000)

fia_plots · table¶

USDA FIA per-plot identification from per-state {ST}_PLOT.csv. One row per FIA plot. Lat/lon are USDA's publicly-released "fuzzed" coordinates (perturbed so exact plot locations cannot be inferred — FIADB P2 User Guide §2.5.1). plot_cn is the FIA-wide unique plot identifier; join targets are TREE/COND/ SEEDLING via PLT_CN. ETL fans out across states.

Column	Type	Null	Description
`plot_cn`	string	no	FIA-wide unique plot control number (PK)
`prev_plot_cn`	string	yes	Previous-cycle plot CN (FK to fia_plots.plot_cn for remeasurement chains)
`inventory_year`	integer	yes	FIA inventory year (INVYR)
`state_fips`	string	no	2-digit FIPS code identifying the primary state (e.g., '06' for California)
`county_fips`	string	yes	5-digit FIPS code for primary county (e.g., '06037' for Los Angeles County)
`plot_number`	integer	yes	USDA plot number within state/county
`plot_status_cd`	integer	yes	1=Sampled-forest, 2=Sampled-nonforest, 3=Nonsampled
`plot_status_name`	string	yes	Human-readable plot status
`lat`	double	yes	Fuzzed plot latitude (decimal degrees, NAD83)
`lon`	double	yes	Fuzzed plot longitude (decimal degrees, NAD83)
`elev_ft`	integer	yes	Plot elevation in feet
`measure_year`	integer	yes	Year the plot was actually measured (MEASYEAR)

fia_tree_grm · table¶

USDA FIA growth-removal-mortality (GRM) annualized estimates by state, inventory year, species, and estimation type. Derived by joining per-state {ST}_TREE_GRM_ESTN.csv against {ST}_TREE.csv on TRE_CN to attach SPCD, then aggregating per-tree ANN_NET_GROWTH/REMOVALS/MORTALITY into per-group totals. Units are determined by ESTN_UNITS for the row's ESTN_TYPE (typically cubic feet for volume estimates and square feet for basal area). ETL fans out across states.

Column	Type	Null	Description
`state_fips`	string	no	2-digit FIPS code identifying the primary state (e.g., '06' for California)
`inventory_year`	integer	no	FIA inventory year (INVYR)
`species_code`	integer	no	FIA species code (SPCD) — joins to REF_SPECIES upstream
`estimation_type`	string	yes	ESTN_TYPE label (e.g., AL_FOREST, GS_TIMBER) — defines basis + units
`ann_net_growth`	double	yes	Sum of per-tree ANN_NET_GROWTH for the group, in ESTN_UNITS
`annual_removals`	double	yes	Sum of per-tree REMOVALS for the group, in ESTN_UNITS
`annual_mortality`	double	yes	Sum of per-tree MORTALITY for the group, in ESTN_UNITS
`tree_count`	long	yes	Number of tree records contributing to this aggregate

fia_seedlings · table¶

USDA FIA seedling regeneration estimates by state, inventory year, and species, derived from per-state {ST}_SEEDLING.csv. seedlings_per_acre is the sum of TPA_UNADJ across plot-conditions; seedling_count is the sum of microplot TREECOUNT observations. Pairs with forest_metrics for tree+regeneration analytics.

Column	Type	Null	Description
`state_fips`	string	no	2-digit FIPS code identifying the primary state (e.g., '06' for California)
`inventory_year`	integer	no	FIA inventory year (INVYR)
`species_code`	integer	no	FIA species code (SPCD)
`seedling_count`	long	yes	Sum of microplot seedling counts (TREECOUNT)
`seedlings_per_acre`	double	yes	Sum of TPA_UNADJ across plot-conditions for the group
`record_count`	long	yes	Number of SEEDLING rows contributing to this aggregate

fia_down_woody_debris · table¶

USDA FIA down-woody-material biomass aggregated by state, inventory year, and debris category (coarse_woody, fine_woody, duff_litter). Derived from per-state {ST}_COND_DWM_CALC.csv. tons_per_acre is the condition-area weighted mean of the FIA per-condition adjusted dry-biomass tonnage for the category.

Column	Type	Null	Description
`state_fips`	string	no	2-digit FIPS code identifying the primary state (e.g., '06' for California)
`inventory_year`	integer	no	FIA inventory year (INVYR)
`debris_category`	string	no	One of coarse_woody, fine_woody, duff_litter
`tons_per_acre`	double	yes	Condition-area weighted mean of category dry-biomass tons/acre
`record_count`	long	yes	Number of COND_DWM_CALC rows contributing to this aggregate

fia_invasives · table¶

USDA FIA invasive plant species observations aggregated by state, inventory year, and species code, derived from per-state {ST}_INVASIVE_SUBPLOT_SPP.csv. cover_pct_avg is the mean COVER_PCT across reporting subplots. Not every state samples invasives every cycle.

Column	Type	Null	Description
`state_fips`	string	no	2-digit FIPS code identifying the primary state (e.g., '06' for California)
`inventory_year`	integer	no	FIA inventory year (INVYR)
`species_code`	string	no	FIA invasive species code (VEG_SPCD — USDA PLANTS symbol, alphanumeric)
`cover_pct_avg`	double	yes	Average COVER_PCT across reporting subplots
`subplot_count`	long	yes	Number of subplots reporting this species

fia_pop_stratum · table¶

FIA stratification stratum — carries EXPNS (the expansion factor, acres represented per plot) and the ADJ_FACTOR_* adjustments. Join fia_pop_plot_stratum_assgn to attach a plot to its stratum, then weight by EXPNS to turn plot observations into design-based population totals. Without this table only per-plot averages are possible, never state or regional estimates. Carried at its published grain (one row per source row) from per-state {ST}_POP_STRATUM.csv, because aggregating it would defeat its purpose.

Column	Type	Null	Description
`state_fips`	string	no	2-digit FIPS code identifying the primary state (e.g., '06' for California)
`cn`	string	yes	FIA control number (surrogate key), carried as a string because FIA CNs exceed 64-bit range
`estn_unit_cn`	string	yes	Estimation-unit control number — FK to fia_pop_estn_unit.cn
`rscd`	integer	yes
`evalid`	integer	yes	Evaluation identifier (state + year + evaluation type)
`estn_unit`	integer	yes
`stratumcd`	integer	yes
`stratum_descr`	string	yes
`statecd`	integer	yes	FIA state code
`p1pointcnt`	integer	yes
`p2pointcnt`	integer	yes
`expns`	double	yes	Expansion factor — acres represented by each plot in this stratum. Multiply plot values by this to get population totals.
`adj_factor_macr`	double	yes	Adjustment factor for macroplot-based attributes
`adj_factor_subp`	double	yes	Adjustment factor for subplot-based attributes
`adj_factor_micr`	double	yes	Adjustment factor for microplot-based attributes
`adj_factor_cwd`	double	yes
`adj_factor_fwd_sm`	double	yes
`adj_factor_fwd_lg`	double	yes
`adj_factor_duff`	double	yes
`created_date`	string	yes
`modified_date`	string	yes
`adj_factor_pile`	double	yes
`adj_factor_regen_micr`	string	yes
`adj_factor_inv_subp`	string	yes
`adj_factor_p2veg_subp`	string	yes
`adj_factor_grndlyr_microquad`	string	yes
`adj_factor_soil`	string	yes

fia_pop_plot_stratum_assgn · table¶

Assigns each plot (PLT_CN) to a stratum (STRATUM_CN) within an evaluation (EVALID). This is the join key between plot-level observations and the stratification that expands them to population totals. Carried at its published grain (one row per source row) from per-state {ST}_POP_PLOT_STRATUM_ASSGN.csv, because aggregating it would defeat its purpose.

Column	Type	Null	Description
`state_fips`	string	no	2-digit FIPS code identifying the primary state (e.g., '06' for California)
`cn`	string	yes	FIA control number (surrogate key), carried as a string because FIA CNs exceed 64-bit range
`stratum_cn`	string	yes	Stratum control number — FK to fia_pop_stratum.cn
`plt_cn`	string	yes	Plot control number — FK to fia_plots
`statecd`	integer	yes	FIA state code
`invyr`	integer	yes
`unitcd`	integer	yes
`countycd`	integer	yes
`plot`	integer	yes
`rscd`	integer	yes
`evalid`	integer	yes	Evaluation identifier (state + year + evaluation type)
`estn_unit`	integer	yes
`stratumcd`	integer	yes
`created_date`	string	yes
`modified_date`	string	yes

fia_pop_estn_unit · table¶

Estimation unit — the geographic partition a stratification is built within, carrying its land and total area. The denominator for area-based estimates. Carried at its published grain (one row per source row) from per-state {ST}_POP_ESTN_UNIT.csv, because aggregating it would defeat its purpose.

Column	Type	Null	Description
`state_fips`	string	no	2-digit FIPS code identifying the primary state (e.g., '06' for California)
`cn`	string	yes	FIA control number (surrogate key), carried as a string because FIA CNs exceed 64-bit range
`eval_cn`	string	yes	Evaluation control number — FK to fia_pop_evaluations
`rscd`	integer	yes
`evalid`	integer	yes	Evaluation identifier (state + year + evaluation type)
`estn_unit`	integer	yes
`estn_unit_descr`	string	yes
`statecd`	integer	yes	FIA state code
`arealand_eu`	double	yes
`areatot_eu`	double	yes
`area_used`	double	yes
`area_source`	string	yes
`p1pntcnt_eu`	integer	yes
`p1source`	integer	yes
`created_date`	string	yes
`modified_date`	string	yes

fia_pop_eval_grp · table¶

Evaluation group: the set of evaluations that together form one reportable inventory cycle for a state (EVAL_GRP, e.g. 'AK2023'). Carried at its published grain (one row per source row) from per-state {ST}_POP_EVAL_GRP.csv, because aggregating it would defeat its purpose.

Column	Type	Null	Description
`state_fips`	string	no	2-digit FIPS code identifying the primary state (e.g., '06' for California)
`cn`	string	yes	FIA control number (surrogate key), carried as a string because FIA CNs exceed 64-bit range
`rscd`	integer	yes
`eval_grp`	integer	yes
`eval_grp_descr`	string	yes
`statecd`	integer	yes	FIA state code
`notes`	string	yes
`created_date`	string	yes
`modified_date`	string	yes

fia_pop_eval_attribute · table¶

Maps an evaluation to the attributes it supports, so a consumer can tell which estimates a given EVALID is valid for. Carried at its published grain (one row per source row) from per- state {ST}_POP_EVAL_ATTRIBUTE.csv, because aggregating it would defeat its purpose.

Column	Type	Null	Description
`state_fips`	string	no	2-digit FIPS code identifying the primary state (e.g., '06' for California)
`cn`	string	yes	FIA control number (surrogate key), carried as a string because FIA CNs exceed 64-bit range
`eval_cn`	string	yes	Evaluation control number — FK to fia_pop_evaluations
`attribute_nbr`	integer	yes
`statecd`	integer	yes	FIA state code
`created_date`	string	yes
`modified_date`	string	yes

fia_pop_evaluations · table¶

USDA FIA population evaluations — the statistically-valid samples used to publish state estimates. Joins {ST}_POP_EVAL.csv with {ST}_POP_EVAL_TYP.csv on EVAL_CN to emit one row per (evaluation, evaluation type) pair. eval_type values include EXPVOL (volume), EXPCURR (current area), EXPGROW, EXPMORT, EXPREMV, etc.

Column	Type	Null	Description
`state_fips`	string	no	2-digit FIPS code identifying the primary state (e.g., '06' for California)
`eval_cn`	string	no	POP_EVAL.CN — PK for the evaluation record
`eval_grp_cn`	string	yes	POP_EVAL_GRP foreign key
`eval_descr`	string	yes	Human-readable description of the evaluation
`start_invyr`	integer	yes	First inventory year covered by the evaluation
`end_invyr`	integer	yes	Last inventory year covered by the evaluation
`eval_type`	string	yes	Evaluation type from POP_EVAL_TYP.EVAL_TYP (EXPVOL, EXPCURR, ...)

nps_units · table¶

National Park Service unit boundaries — parks, monuments, recreation areas, seashores, parkways, and all other NPS-designated units. ~430 units. Reference and dimension table; FK target for nps_visitation.unit_code. Joins to geo.states and geo.counties for gateway-community economic impact analysis. Updated annually in April following annual report administrative updates. geometry_wkt carries the unit's own WGS84 boundary polygon, sourced from the same ArcGIS query at ingest time; this table still carries no county_fips column of its own (see nps_visitation's comment), so a county-grain join needs a spatial predicate against geometry_wkt (e.g. DuckDB ST_Intersects with geo.counties), following the pattern lands.county_federal_fee_land_share already uses for padus_federal_fee_lands.

Column	Type	Null	Description
`unit_code`	string	no	NPS 4-letter unit code (e.g., 'YOSE', 'GRCA')
`unit_name`	string	yes	Full unit name
`unit_type`	string	yes	National Parks / Monuments / Recreation Areas / Seashores / etc.
`state_abbr`	string	yes	2-character state abbreviation (e.g., 'CA', 'WY')
`region`	string	yes	NPS region (e.g., 'Pacific West', 'Northeast')
`gross_acres`	double	yes	Approximate unit area in acres, derived from the ArcGIS Shape__Area attribute (always returned in the layer's Web Mercator storage SR regardless of outSR) with a cos(lat)^2 correction for Mercator's latitude-dependent area inflation, using the geometry's mean vertex latitude as the representative point.
`geometry_wkt`	string	yes	Simplified boundary WKT polygon (for display only)

nps_visitation · table¶

National Park Service monthly visitation statistics from the IRMA Statistics API. One row per unit_code × visit_year × visit_month (no annual summary rows), with recreation_visits and non_recreation_visits as the two visit-count metrics. Joins to nps_units via unit_code. See lands_nps_gateway_impact and lands_nps_gateway_employment for pre-built annual aggregations per unit; note nps_units carries no county_fips, so county-level employment correlation is not directly supported.

Column	Type	Null	Description
`unit_code`	string	no	FK to nps_units (NPS 4-letter code)
`visit_year`	integer	no	Visit year (partition key)
`visit_month`	integer	no	Month (1–12); all rows are monthly (no annual summary rows)
`recreation_visits`	integer	yes	Recreational visitor count
`non_recreation_visits`	integer	yes	Non-recreational visits (commuters, government, etc.)

blm_field_offices · table¶

Bureau of Land Management administrative unit boundaries — field offices and district offices that manage federal subsurface mineral rights, grazing allotments, and surface use permits. ~150 field offices. Primarily useful as a geographic reference for joining extractive industry activity to SEC filings and FEC oil/gas sector donor geography. Updated annually in June. geometry_wkt is the office's own point location only (this table's source layer carries point geometry, not the administrative boundary polygon — see the column comment); for a federal-fee land boundary/ownership polygon to spatially join against counties or fire perimeters, use lands.padus_federal_fee_lands instead.

Column	Type	Null	Description
`office_code`	string	no	BLM administrative unit code (e.g., 'UTMO')
`office_name`	string	yes	Office name (e.g., 'Moab Field Office')
`office_type`	string	yes	Field Office / District Office / State Office
`state_abbr`	string	yes	2-character state abbreviation (e.g., 'UT', 'WY')
`geometry_wkt`	string	yes	Office location as WGS84 WKT POINT — this FeatureServer layer (0) carries only the office's point location, not its administrative boundary polygon (that lives on separate per-level layers of the same service: 1=state, 2=district, 3=field, 4=other, joinable by the same office_code/ADM_UNIT_CD but not fetched here).

onrr_revenues · table¶

Office of Natural Resources Revenue (ONRR) royalty payments for oil, gas, coal, and other minerals extracted from federal and tribal lands, full history from FY 2004 in a single bulk download. Row grain is every descriptive column together (fiscal_year, land_class, land_category, state_name, county_name, county_fips, offshore_region, revenue_type, mineral_lease_type, commodity, product) — revenue is the only measure. county_fips alone does not identify a row: ONRR carries multiple distinct rows sharing one FIPS (e.g. "Kenai (Pre)" vs "Kenai (Post)" county_name), and offshore rows (null state_name/county_name/county_fips) are distinguished by offshore_region instead. Primary financial link between BLM-managed subsurface rights and SEC 10-K disclosures for E&P and mining companies. Source: revenuedata.onrr.gov (domain moved from revenuedata.doi.gov in 2024).

Column	Type	Null	Description
`fiscal_year`	integer	no	Fiscal year of revenue collection (FY 2004–present)
`land_class`	string	yes	Federal / Native American / Federal Offshore
`land_category`	string	yes	Onshore / Offshore / Gulf of Mexico / Pacific
`state_name`	string	yes	State name (e.g., 'Wyoming'); null for offshore rows
`county_name`	string	yes	County name; null for offshore/tribal rows
`county_fips`	string	yes	5-digit FIPS (null for offshore/tribal)
`offshore_region`	string	yes	Offshore region name (Gulf of Mexico, Pacific, etc.); null for onshore
`revenue_type`	string	yes	Royalties / Bonus / Rents / Other Revenues
`mineral_lease_type`	string	yes	Oil & Gas / Coal / Geothermal / Hardrock / etc.
`commodity`	string	yes	Oil / Gas / Coal / Geothermal / Renewables / Other
`product`	string	yes	Specific product (e.g., 'Crude Oil', 'Coalbed Methane')
`revenue`	double	yes	Revenue collected (USD)

pilt_county_payments · table¶

DOI Payments in Lieu of Taxes (PILT) — federal payments to local governments for tax-exempt federal land within their boundaries, one row per (fiscal_year, state_code, area_name). Counterpart to onrr_revenues (federal mineral-revenue sharing): PILT compensates for the property-tax base federal land removes, independent of resource extraction. area_type is derived from the source label's own COUNTY/PARISH/BOROUGH/TOWN/... wording (null when the label carries none, e.g. DC/Guam/the Virgin Islands, each a single territory-wide row) rather than assumed from the state, since some states mix area types (Vermont's table has both towns and county-level subtotal rows). No bulk export exists; one fetch per (fiscal_yr, state_code) returns that state's full county list. Source: pilt.doi.gov/counties.cfm.

Column	Type	Null	Description
`fiscal_year`	integer	no	Federal fiscal year of the payment (FY1999-present; PK component)
`state_code`	string	no	USPS state/territory code fetched (PK component)
`area_name`	string	no	County/parish/borough/town/territory label as published (PK component)
`area_type`	string	yes	COUNTY/PARISH/BOROUGH/TOWN/CITY/... parsed from area_name; null when unrecognized
`county_fips`	string	yes	5-digit county FIPS, resolved against the Census county-name crosswalk (D-282; same lookup energy.eia_power_plants/banking.history use). NULL for roughly 7% of FY2025-scale payment dollars — New England towns/plantations below the county level, independent cities and territory-wide rows (DC/Guam/PR/VI) the crosswalk has no county-equivalent entry for at all. This is a real gap in standard county geography for those area types, not a lookup failure to fix by guessing.
`payment_dollars`	bigint	yes	PILT payment for the area (USD)
`total_acres`	bigint	yes	Total federal acres in the area used to compute the payment

va_facilities · table¶

Veterans Health Administration facility roster — VA medical centers (VAMC), community-based outpatient clinics (CBOC), Vet Centers (VCTR), VA nursing homes (VANH), and residential rehabilitation treatment programs (RRTP), geocoded. One row per facility (1,046 confirmed live). Answers "nearest VA facility to a given location" directly via latitude/longitude — the corpus previously had no way to do this, only a rurality proxy via census.acs_veterans joined to health.ahrf_physician_supply's RUCC. Joins to geo.counties via county_fips and geo.states via state_abbr. Source: HIFLD Veterans Health Administration Medical Facilities ArcGIS FeatureServer.

Column	Type	Null	Description
`station_number`	string	no	VA station number (PK)
`facility_name`	string	yes	Facility name
`street_address`	string	yes	Street address (line 1)
`street_address2`	string	yes	Street address line 2 (suite/building), when present
`city`	string	yes	City
`state_fips`	string	yes	2-digit FIPS code identifying the primary state (e.g., '06' for California)
`state_abbr`	string	yes	2-character USPS state abbreviation (FK to geo.state_ref.state_abbr)
`zip`	string	yes	ZIP code
`county_name`	string	yes	County name, as published by HIFLD
`county_fips`	string	yes	5-digit FIPS code for primary county (e.g., '06037' for Los Angeles County)
`latitude`	double	yes	Facility latitude
`longitude`	double	yes	Facility longitude
`facility_type`	string	yes	HIFLD's primary-service classification (PRIM_SVC): VAMC (medical center), CBOC (community-based outpatient clinic), VCTR (Vet Center), VANH (VA nursing home), RRTP (residential rehabilitation treatment program), or IOC (independent/other outpatient clinic — HIFLD's own category; does not always align with the source's per-type Y/N flags, e.g. a facility marked RRTP=YES can still carry PRIM_SVC=IOC as its primary classification). Use this column, not a derived flag, to classify a facility.
`naics_code`	string	yes	NAICS industry code
`naics_description`	string	yes	NAICS industry description
`visn`	string	yes	Veterans Integrated Service Network number
`congressional_district`	string	yes	115th Congress congressional district code (e.g. 'CD115_AZ_0404')
`phone`	string	yes	Facility phone number

padus_federal_fee_lands · table¶

USGS GAP Analysis Project's authoritative federal-fee-land layer (Federal_Fee_Managers_Authoritative_PADUS, part of PAD-US) — one row per federal-fee-land polygon (~5,361 features) with owning/managing agency (own_type/own_name, mang_type/mang_name), designation_type, state_abbr, acres, and a simplified WGS84 boundary polygon (geometry_wkt). Federal-fee ownership only (not the full multi-hundred-thousand-polygon PAD-US inventory across all ownership classes — state/tribal/local/NGO/private land is out of scope here). No county_fips column; assigning per-county ownership shares requires a spatial join (DuckDB ST_Intersects) against geo.counties, which is a natural follow-up view, not done at ingest time. Snapshot table with no year partition (refreshed via overwritePartitions). Answers the per-county, cross-agency federal-fee-ownership-share need behind D-054 in one place — own_type/mang_name lets a query filter or compare ownership across USFS, BLM, NPS, and other federal managers without joining three separately-sourced tables. national_forests and nps_units now also carry their own geometry_wkt (their unit's own boundary polygon, captured directly at ingest); blm_field_offices carries only a point location (see its column comment). Also gives disasters.wildfire_perimeters a federal-fee-land layer to compute federal-vs-nonfederal wildfire exposure against (D-176). Source: Federal_Fee_Managers_Authoritative_PADUS/FeatureServer/0 (official USGS GAP Analysis Project ArcGIS org, item v01gqwM5QqNysAAi). WARNING: do not SUM(acres) for a total federal-fee acreage figure — this layer's own rows are not mutually exclusive despite its "Authoritative" name. Verified live 2026-09-12 directly against the source FeatureServer (not an artifact of this table's ingestion — our own SUM(acres) matches the source's own aggregate query exactly): summing all 5,361 rows gives ~2.25B acres, near the entire land area of the US; even filtering to category='Fee' (ownership, excluding Proclamation boundaries) still gives ~2.10B acres. Grouping by mang_name shows BLM/USFS/NPS/USACE/USBR sums are each close to their independently known real totals (e.g. USFS ~193M acres, matching the National Forest System almost exactly) — the inflation is concentrated in FWS, whose rows include near-duplicate polygon pairs for the same unit at nearly identical (not exactly equal) acreage, e.g. Pacific Remote Islands Marine National Monument appears twice (310,260,466 and 310,260,380 acres), Papahanaumokuakea Marine National Monument three times (283,392,764 / 283,392,311 / 88,143,702 for a "Monument Original" sub-record), Mariana Trench NWR twice (50,559,347 / 50,559,316), Arctic NWR twice (19,637,182 / 19,637,012). This is a source data-quality defect in PAD-US's own authoritative layer, not fixable by de-duplicating on any single column here. A query needing a real distinct-area total should use CRS's published federal land figure (~640-650M acres) rather than SUM(acres) on this table, or compute a true union area via ST_Union/ST_Area over geometry_wkt (costly but correct). This table remains the right source for per-parcel attribute lookups, agency/designation filters, and spatial joins (e.g. against geo.counties or disasters.wildfire_perimeters) where exact row-for-row overlap doesn't matter.

Column	Type	Null	Description
`padus_id`	string	no	Surrogate PK derived from the service's OBJECTID (PADUS-)
`own_type`	string	yes	Owner type code (FED/TRIB/STAT/DIST/LOC/NGO/PVT/JNT/UNK/TERR/DESG)
`own_name`	string	yes	Owning agency or entity name (e.g., 'Bureau of Land Management')
`mang_type`	string	yes	Managing entity type code (general depiction of federal land manager)
`mang_name`	string	yes	Managing agency name (e.g., 'BLM', 'USFS', 'DOD')
`designation_type`	string	yes	Designation/protection category (e.g., 'MIL', 'NWR', 'NF')
`unit_name`	string	yes	Named unit (e.g., 'Yosemite National Park')
`state_abbr`	string	yes	2-character state abbreviation
`acres`	double	yes	GIS-calculated parcel area in acres
`geometry_wkt`	string	yes	Simplified boundary WKT polygon (for display only)

lands_forest_condition_metrics · view¶

Joins forest_inventory (basal area from COND CSV) with forest_metrics (live-tree volume and carbon from TREE CSV) on the shared composite key. Provides a single row per state × forest type group × ownership class × inventory year with all four FIA-derived metrics (basal_area_sqft, trees_per_acre, live_volume_cuft, carbon_stock_tons). Use for carbon-stock trend analysis, timber-volume benchmarking, and cross-schema joins to sec.filing_metadata or econ.fred_indicators.