gcs-landbrugsareal-data: Скилл для анализа данных по сельхозу

--- name: gcs-landbrugsareal-data description: | Activates when querying agricultural land and field data from GCS. Use this skill for: field boundaries, crop data, land use, organic farming, production estimates, cadastral data, agricultural blocks, building data. Keywords: landbrugsareal, field, mark, marker, crop, afgrøde, organic, økologisk, production, areal, cadastral, matrikel --- # GCS Landbrugsareal (Agricultural Land) Data Catalog Field boundaries, crop data, land use, and agricultural production data. ## Frontend Metrics Supported | Metric Key | Danish Name | Description | |------------|-------------|-------------| | `land_use` | Arealanvendelse | Land use categories | | `organic_farming` | Økologisk areal | Organic farming area | | `production` | Afgrødeproduktion | Crop production estimates | | `crop_diversity` | Afgrødediversitet | Diversity of crops grown | ## Available Datasets ### Silver Layer #### FVM Marker - Field Boundaries (617K rows/year) **Path**: `gs://$GCS_BUCKET/silver/fvm_marker_{year}/*/data.parquet` **Years Available**: 2008-2025 | Column | Type | Description | Example | |--------|------|-------------|---------| | field_id | string | Unique field identifier | 610341-27-1-0 | | area_ha | float | Field area in hectares | 12.45 | | cvr_number | string | Company CVR (8 digits) | 31373077 | | crop_code | int | Crop type code | 1 | | crop_name | string | Crop type name | Vårbyg | | block_id | string | Agricultural block ID | 610341-27 | | geometry | binary | Field polygon (WKB, EPSG:4326) | - | | year | int | Data year | 2024 | | field_uuid | string | UUID for field | - | | municipality | string | Municipality code | 0101 | | is_organic | bool | Organic certification | true | | organic_conversion_date | date | Date of organic conversion | 2020-01-15 | **Schema (introspected)**: ``` field_id: string area_ha: double cvr_number: string crop_code: int64 crop_name: string block_id: string geometry: binary (WKB) year: int64 field_uuid: string municipality: string is_organic: bool organic_conversion_date: date32 [20 columns total] ``` #### Agricultural Blocks **Path**: `gs://$GCS_BUCKET/silver/agricultural_blocks_{year}/*/data.parquet` | Column | Type | Description | |--------|------|-------------| | block_id | string | Block identifier | | geometry | binary | Block polygon (WKB) | | total_area_ha | float | Total block area | #### Cadastral (2.16M rows) **Path**: `gs://$GCS_BUCKET/silver/cadastral/*/data.parquet` | Column | Type | Description | Example | |--------|------|-------------|---------| | bfe_number | int64 | Cadastral parcel ID | 1234567 | | business_event | string | Last registration event | - | | registration_from | timestamp | Registration date | - | | is_worker_housing | bool | Worker housing flag | false | | is_common_lot | bool | Common lot flag | false | | agricultural_notation | string | Agricultural notation | - | | geometry | binary | Parcel polygon (WKB) | - | #### BBR Buildings (579K rows) **Path**: `gs://$GCS_BUCKET/silver/bbr_buildings/*/joined_buildings.parquet` | Column | Type | Description | |--------|------|-------------| | building_uuid | string | Unique building ID | | geo_building_polygon | binary | Building footprint (WKB) | | geo_building_centroid | binary | Building center point | | building_type | string | Building type code | | building_floor_area_sqm | float | Floor area in m² | | building_usage_category | string | Usage category | | current_use | string | Current use description | | address | string | Building address | #### DAGI Kommuner (99 rows) **Path**: `gs://$GCS_BUCKET/silver/dagi_kommuner/*/data.parquet` | Column | Type | Description | |--------|------|-------------| | code | string | Municipality code | | name | string | Municipality name | | region_code | string | Region code | | geometry | binary | Municipality polygon (WKB) | | area_m2 | float | Area in m² | | centroid_x | float | Centroid longitude | | centroid_y | float | Centroid latitude | ### Gold Layer #### Field Production (617K rows/year) **Path**: `gs://$GCS_BUCKET/gold/field_production_{year}/*/data.parquet` | Column | Type | Description | Example | |--------|------|-------------|---------| | field_id | string | Field identifier | 610341-27-1-0 | | block_id | string | Block identifier | 610341-27 | | cvr_number | string | Company CVR | 31373077 | | year | int | Production year | 2024 | | area_ha | float | Field area | 12.45 | | crop_type | string | Crop type name | Vårbyg | | organic_farming | bool | Organic flag | true | | landsdel_code | string | Region code | - | | landsdel_name | string | Region name | Nordjylland | | dst_regions | string | DST region classification | - | | kommune_name | string | Municipality name | København | | yield_estimate_hkg_ha | float | Yield estimate hkg/ha | 52.3 | | production_estimate_hkg | float | Total production hkg | 651.14 | | field_uuid | string | UUID | - | | primary_field_id | string | Primary field reference | - | **Schema (introspected)**: ``` field_id: string block_id: string cvr_number: string year: int64 area_ha: double crop_type: string organic_farming: bool landsdel_code: string landsdel_name: string dst_regions: string kommune_name: string yield_estimate_hkg_ha: double production_estimate_hkg: double field_uuid: string primary_field_id: string [18 columns total] ``` ## Common Queries ### Get All Fields for a CVR ```python import pyarrow.parquet as pq from google.cloud import storage import io client = storage.Client() bucket = client.bucket('$GCS_BUCKET') # Read FVM marker data for 2024 blob = bucket.blob('silver/fvm_marker_2024/2025-01-10/data.parquet') buffer = io.BytesIO() blob.download_to_file(buffer) buffer.seek(0) df = pq.read_table(buffer).to_pandas() # Filter by CVR company_fields = df[df['cvr_number'] == '31373077'] print(f"Total fields: {len(company_fields)}") print(f"Total area: {company_fields['area_ha'].sum():.2f} ha") ``` ### Calculate Organic Farming Percentage by Municipality ```python # Group by municipality and organic status organic_stats = df.groupby(['municipality', 'is_organic']).agg({ 'area_ha': 'sum' }).unstack(fill_value=0) organic_stats.columns = ['conventional', 'organic'] organic_stats['organic_pct'] = ( organic_stats['organic'] / (organic_stats['conventional'] + organic_stats['organic']) * 100 ) organic_stats = organic_stats.sort_values('organic_pct', ascending=False) ``` ### Get Production Estimates for a Crop ```python # Read gold production data blob = bucket.blob('gold/field_production_2024/2025-01-10/data.parquet') buffer = io.BytesIO() blob.download_to_file(buffer) buffer.seek(0) production = pq.read_table(buffer).to_pandas() # Filter for wheat wheat = production[production['crop_type'].str.contains('hvede', case=False)] print(f"Total wheat area: {wheat['area_ha'].sum():.2f} ha") print(f"Average yield: {wheat['yield_estimate_hkg_ha'].mean():.2f} hkg/ha") print(f"Total production: {wheat['production_estimate_hkg'].sum():.2f} hkg") ``` ### Crop Diversity Analysis ```python # Count unique crops per CVR crop_diversity = df.groupby('cvr_number').agg({ 'crop_name': 'nunique', 'area_ha': 'sum' }).reset_index() crop_diversity.columns = ['cvr_number', 'crop_count', 'total_area'] crop_diversity = crop_diversity.sort_values('crop_count', ascending=False) ``` ### Read Geometry with GeoPandas ```python import geopandas as gpd import io # Read with geometry blob = bucket.blob('silver/fvm_marker_2024/2025-01-10/data.parquet') buffer = io.BytesIO() blob.download_to_file(buffer) buffer.seek(0) # Read as GeoDataFrame gdf = gpd.read_parquet(buffer) gdf = gdf.set_crs('EPSG:4326') # Set CRS if not set # Plot a sample gdf[gdf['cvr_number'] == '31373077'].plot() ``` ## Join Keys | This Dataset | Join Column | Target Dataset | Target Column | |--------------|-------------|----------------|---------------| | fvm_marker | cvr_number | subsidies | cvr_number | | fvm_marker | field_id | field_production | field_id | | fvm_marker | block_id | agricultural_blocks | block_id | | fvm_marker | field_uuid | pesticide_disaggregation | field_uuid | | fvm_marker | municipality | dagi_kommuner | code | | field_production | field_id | nles5_nitrogen | field_id | | cadastral | bfe_number | property_owners | bestemtFastEjendomBFENr | ## Data Quality Notes ### FVM Marker - **Update frequency**: Annual (new year data released ~Q2) - **Coverage**: All registered agricultural fields in Denmark - **Years available**: 2008-2025 (varies by data source) - **Geometry**: WKB format, EPSG:4326 ### Field Production - **Update frequency**: Annual after harvest statistics - **Coverage**: Production estimates for all registered fields - **Caveat**: Yield estimates are modeled, not measured ### Cadastral - **Update frequency**: Weekly - **Coverage**: All cadastral parcels in Denmark - **Note**: Use for property boundaries, not field boundaries ## Related Skills - **okonomi/** - Subsidies linked to fields via CVR and markbloknummer - **miljo/** - Environmental data at field level (pesticides, nitrogen) - **husdyr/** - Livestock density per land area ## GCS Paths Reference ```bash # List available FVM marker years gsutil ls gs://$GCS_BUCKET/silver/ | grep fvm_marker # List field production years gsutil ls gs://$GCS_BUCKET/gold/ | grep field_production # List cadastral snapshots gsutil ls gs://$GCS_BUCKET/silver/cadastral/ # List DAGI municipality data gsutil ls gs://$GCS_BUCKET/silver/dagi_kommuner/ ``` ## Yearly Data Pattern For datasets with yearly data, use this pattern: ```python # Query specific year year = 2024 path = f'silver/fvm_marker_{year}/*/data.parquet' # Query all years (use with caution - large data) path = 'silver/fvm_marker_*/*/data.parquet' ```