Skip to content

Modeling Conservation Agriculture

CropForge v0.7.0 introduces Topographical Physics, enabling the simulation of land preparation effects on water dynamics, soil erosion, and plant growth in a true 3D spatial context.

This tutorial walks through the conservation_ag_trial.py example script, demonstrating how to set up a comparison between Conventional Tillage and Conservation Agriculture (Contour Bunding) on a sloped terrain.

The Goal

We want to quantitatively demonstrate how contour bunds (a conservation agriculture technique) reduce soil erosion compared to conventional tillage on an undulating, sloped field under heavy rainfall.

Step 1: Terrain Generation

First, we generate a procedural undulating terrain. This provides the slope necessary to drive runoff and erosion.

from cropforge.terrain import Terrain

# The terrain has a built-in slope/undulation to drive runoff and erosion
terrain = Terrain.procedural(generator="undulating", rows=30, cols=30, resolution_m=1.0)

Step 2: Weather Setup

We simulate 60 days with frequent heavy rainfall events to stress test the erosion mechanics.

import pandas as pd
from cropforge.loaders import Weather

weather_df = pd.DataFrame({
    "day": range(1, 61),
    "doy": range(150, 210),
    "temp_max_c": 32.0,
    "temp_min_c": 22.0,
    "temp_mean_c": 27.0,
    "radiation_mj_m2": 20.0,
    # Heavy rain every 5 days
    "rainfall_mm": [30.0 if i % 5 == 0 else 2.0 for i in range(1, 61)],
    "et0_mm": 5.0,
    "wind_speed_ms": 3.5,
    "humidity_pct": 75.0,
    "co2_ppm": 415.0
})
weather = Weather(weather_df.set_index("day"))

Step 3: Setting Up the Fields

We create two fields side-by-side using the identical terrain and weather, but differing LandPrep configurations.

Field A: Conventional Tillage

Conventional tillage starts with a high surface roughness, but it decays quickly under heavy rain. It offers no physical barriers to lateral flow.

from cropforge.farm import Field
from cropforge.land_prep import ConventionalTill

field_a = Field(name="Conventional Till", rows=30, cols=30)
field_a.set_terrain(terrain)
field_a.set_weather(weather)
field_a.set_crop(maize)
field_a.set_land_prep(ConventionalTill())
farm.add_field(field_a)

Field B: Contour Bunds

Contour bunds are physical barriers placed along the contour lines of the slope. They trap water, reducing runoff velocity and minimizing erosion.

from cropforge.land_prep import ContourBund

field_b = Field(name="Contour Bunds", rows=30, cols=30)
field_b.set_terrain(terrain)
field_b.set_weather(weather)
field_b.set_crop(maize)
# Bunds trap water, reducing runoff velocity and erosion
field_b.set_land_prep(ContourBund(bund_height_m=0.3, interval_m=1.0))
farm.add_field(field_b)

Step 4: Enabling Topographical Physics

To activate the 3D physics engines, we must explicitly enable them before running the simulation.

farm.use_physics(
    et0=True,
    water_balance=True,              # Required for runoff
    lateral_flow=True,               # D8 runoff routing
    clod_dynamics=True,              # Rain melts the clods over time
    erosion=True,                    # RUSLE-based erosion index
    slope_radiation_correction=True, # Solar incidence adjusted by slope/aspect
    terrain_wind=True,               # Topographical wind field
    root_clamping=True,              # Topographical root constraints
    wind_direction_deg=270.0         # Wind coming from the West
)

Step 5: Run and Visualize

Run the simulation: 

python examples/conservation_ag_trial.py

Then, launch the CropForge Dashboard: 

cropforge-dash

Open your browser to http://localhost:8050. 1. Select "Field: Conventional Till". 2. Scrub to Day 60. 3. Click "Open 3D Terrain View". 4. In the Surface Overlay Variable dropdown, select Cumulative Erosion Index. 5. Observe the high erosion channels forming on the steep sections of the terrain. 6. Switch to "Field: Contour Bunds" and observe how the bunds physically block the flow paths, significantly lowering the overall erosion index.