I built Land Atlas, a web application for analyzing the agricultural potential of land listings.
The main goal is to answer:
What is the quality of the soil?
How farmable is the property?
Which crops are best suited to its soil and growing conditions?
Most property listings contain price, acreage, photographs, and a short description, but provide little structured information about agricultural viability.
Land Atlas combines listing and geospatial data with:
Soil type and texture
Soil productivity indicators
pH and organic matter
Drainage and water capacity
Slope and terrain
Flooding and ponding
Climate and growing conditions
Crop requirements
The application attempts to convert those inputs into:
Soil-quality assessments
Farmability scores
Property limitations
Crop-suitability rankings
One difficult part has been accurately resolving rural listings. Postal cities, listing labels, coordinates, and physical parcel locations frequently disagree.
I recently changed the location hierarchy to prioritize parcel identifiers and listing-provided coordinates over ordinary address geocoding.
I would appreciate feedback on:
Soil-scoring methodology
Crop-suitability modeling
Parcel-level versus point-level analysis
Communicating uncertainty
Global agricultural datasets
Normalizing data between jurisdictions
Demo: https://land-atlas-production.up.railway.app/
I built Land Atlas, a web application for analyzing the agricultural potential of land listings. The main goal is to answer: What is the quality of the soil? How farmable is the property? Which crops are best suited to its soil and growing conditions? Most property listings contain price, acreage, photographs, and a short description, but provide little structured information about agricultural viability. Land Atlas combines listing and geospatial data with: Soil type and texture Soil productivity indicators pH and organic matter Drainage and water capacity Slope and terrain Flooding and ponding Climate and growing conditions Crop requirements The application attempts to convert those inputs into: Soil-quality assessments Farmability scores Property limitations Crop-suitability rankings One difficult part has been accurately resolving rural listings. Postal cities, listing labels, coordinates, and physical parcel locations frequently disagree. I recently changed the location hierarchy to prioritize parcel identifiers and listing-provided coordinates over ordinary address geocoding. I would appreciate feedback on: Soil-scoring methodology Crop-suitability modeling Parcel-level versus point-level analysis Communicating uncertainty Global agricultural datasets Normalizing data between jurisdictions Demo: https://land-atlas-production.up.railway.app/