CropForge

Open-source, code-first virtual farm runtime for agricultural researchers.

CropForge lets you define a crop simulation entirely in Python. You write the model equations; CropForge handles time-stepping, spatial state management, logging, and visual playback.

pip install cropforge

What's New in v0.7.0

Topographical Physics

• Solar Incidence Engine: Modifies radiation absorption dynamically based on slope, aspect, and solar declination calculations.
• Wind Shadow Engine: Models localized wind fields based on prevailing direction, offering leeward shelter and ridgeline intensification.
• Clod Dynamics: Exponential decay of soil surface roughness during heavy rainfall events.
• Topographical Erosion Engine: Incorporates a grid-based RUSLE model evaluating slope gradient, daily surface runoff, vegetation cover, and roughness dampening.
• 3D Observable Updates: "Cumulative Erosion Index" and "Surface Runoff" can now be mapped directly onto the WebGL terrain viewport via the Parquet data layer.
• Root Clamping: Strict clamp downward root growth into effective_soil_depth_m, simulating the physical boundaries of carved terraces or deep furrows.

What's New in v0.6.0

3D Topography and Land Preparation

• Terrain Engine: Supports procedural generation, CSV import, and GeoTIFF ingestion for 3D elevation grids.
• Land Preparation: Pre-simulation agronomic modifiers (RidgeFurrow, ContourBund, Terrace, DeepTillage, ConservationTillage) shape the land and alter soil physics.
• D8 Hydrology Coupling: The physics engine natively routes lateral water and nutrient flows across the modified topography using the D8 steepest-descent algorithm.
• 3D Dashboard Modal: A high-performance WebGL viewport seamlessly transitions from 2D heatmaps to an interactive 3D terrain viewer, letting you map agronomic variables (Nitrogen, Moisture) directly onto physical crop instances.

What's New in v0.5.0

White Minimal Scientific Dashboard

A complete rework of the visualisation layer built on Material Design:

• Sleek, clean white minimal theme with refined borders and ample whitespace.
• Persistent Farm Inspector (always visible, no click-to-reveal).
• Left sidebar with all controls; right sidebar with live charts.
• Animated preloader with per-field progress tracking.
• All panels renamed to researcher-facing terminology.

Official Crop Plugins — StandardWheat & StandardMaize

First-party, PRD-verified crop plugins distributed as cropforge.plugins:

from cropforge.plugins import StandardWheat, StandardMaize

field.use_plugin(StandardWheat)    # CERES-Wheat phenology + RUE biomass + grain fill
field.use_plugin(StandardMaize)    # C4 parameters + root impedance + water stress

• StandardWheat: 6-stage thermal-time phenology, RUE biomass, grain-fill partitioning into plant.custom['grain_biomass_g'].
• StandardMaize: C4 photosynthesis parameters, hard-pan root clamping, water-stress mortality.
• Both plugins are fully isolated per field — no cross-talk when running dual-plot comparisons.

Advanced Physics Engines

Two new opt-in physics engines, mathematically verified against PRD Crucible criteria:

Beer-Lambert Radiation Interception

farm.use_physics(radiation=True, k_extinction=0.45)
# Writes plant.custom['intercepted_par_mj'] for every plant every day

Implements: PAR_int = solar_rad × 0.5 × (1 − e^(−k × LAI)) Crucible verified: LAI=3.0, k=0.45, rad=15 MJ → 5.5557 MJ ±0.001.

Wind-driven Anisotropic Disease Spread

farm.use_physics(
    disease=True,
    disease_foci=[(15, 15)],          # Initial outbreak coordinates
    disease_wind_direction_deg=270.0,  # From West → spreads East
    disease_spread_rate=0.20,
    disease_anisotropy=0.80,
)

A spatially explicit SIR grid model where infection probability is heavily weighted by wind direction. Crucible verified: eastern half has 2.5× more infections than western half when wind blows East.

Quick Start

from cropforge import Farm, Field, Crop, Soil, Weather
from cropforge.plugins import StandardWheat

farm  = Farm(name="MyFarm", location=(28.6, 77.2))
field = Field(name="Plot A", rows=20, cols=30, area_ha=2.4)
field.set_crop(Crop(species="wheat"))
field.set_weather(Weather.from_csv("data/weather.csv", ...))
field.set_soil(Soil.from_csv("data/soil.csv", apply="uniform"))
farm.add_field(field)

field.use_plugin(StandardWheat)
farm.use_physics(radiation=True, disease=True, disease_wind_direction_deg=270.0)

farm.run(days=90)

See examples/wheat_basic_v2.py, examples/maize_dual_plot_v2.py, and examples/disease_outbreak_trial.py for complete working scripts.

What's New in v0.4.0

• Plugin Ecosystem: Extensible architecture for third-party crop models via PyPI.
• Multi-Season Rotations: Preserve soil state between consecutive runs.
• Compare Dashboard: Compare multiple farm configs and export to CSV.
• Spatial Hydrology: D8 lateral surface water routing across gridded fields.

What's New in v0.2.0

• Opt-In Physics: FAO-56 Penman-Monteith ET0 and Root Impedance models.
• Multi-Field Dashboard: Field Selector, comparative time-series, GxE analysis.

Documentation

Full documentation at cropforge.readthedocs.io.

Licence

MIT — Saswat Sundar Rath, ICAR-IARI Jharkhand, 2026