Model Card: Abiotic Stress Model

What it is

Overview

Environmental stresses such as heat, frost, and cold can be major factors limiting crops productivity. These abiotic stresses can cause significant yield losses and contribute to variability in crop performance across different locations and growing seasons. Early detection and accurate quantification of stress conditions are critical for implementing timely management interventions.

The Abiotic Stress API provides real-time monitoring and forecasting capabilities for major environmental stresses affecting agricultural crops. The system delivers both risk assessment for upcoming stress events and severity measurements for current conditions, enabling growers to make data-driven decisions about crop management practices. Users can get actionable insights on stress timing, intensity, and duration, and this information can be used to make decisions such as irrigation scheduling, variety selection, planting decisions, and protective treatments.

While the API provides valuable quantitative data on stress conditions, it does not replace the need for on-ground verification and professional agronomic expertise in crop management decisions.

Inputs and Outputs

Required field inputs

• Field location (latitude and longitude, point)
• Crop

Optional inputs

• For severity insights: prediction time period (i.e. "start" and "end" dates)
• For risk insights: number of years in the past used to fetch historical weather data

The optional inputs can be modified by users for more custom model tuning, otherwise default values will be used.

Outputs

For each crop stress type, the following insights can be computed by the model and returned by the API:

• Daily stress severity index (0-9) based on actual or forecasted weather data

  • Stress severity ranking: 0=no stress / 9=extreme severity

• Aggregated stress risk (percentage), reflecting the frequency of stress events over a historical period and projected on the upcoming season timeline

  • 0%: the stress (above the defined risk threshold) did not occur once in the last [number of years provided in historical time period]
  • 100%: stress (above the defined risk threshold) occurred every year in the last [number of years provided in historical time period]

The current crops stress types available are the following:

• Diurnal heat stress
• Nighttime heat stress
• Frost stress
• Diurnal and nighttime cold stress

How it works

Algorithm principles

The abiotic stress algorithms quantify daily stress intensity using crop-specific physiological thresholds.

Heat Stress (diurnal): Algorithms monitor daytime temperatures against crop-specific upper thresholds; when temperatures exceed these limits, the system calculates stress. Excessive heat disrupts photosynthesis, impairs pollination, and reduces grain filling, with severity increasing above the threshold temperatures.

Heat Stress (nocturnal): Nighttime temperature algorithms assess disruption to plant respiration and metabolic processes. Elevated night temperatures prevent proper recovery from daytime stress and reduce carbohydrate translocation to developing organs, leading to decreased biomass accumulation and yield potential.

Cold Stress (diurnal and nocturnal): The system evaluates temperatures below optimal ranges but above the freezing thresholds. Suboptimal temperatures slow metabolic processes, delay development, and reduce photosynthetic efficiency.

Frost Stress: Algorithms detect temperatures near or below crop-specific freezing tolerance thresholds. Frost stress damage can result in growth impairment or permanent tissue death.

Model parameters

The model equations use specific parameters related to plant physiology and risk evaluation. These parameters enable optimizing the predictions at crop and region level. The parameters are managed independently of API usage by the agronomy experts for each crop and region.

How to use

Intended use

Pre-season Planning (Risk Assessment): Abiotic stress risk forecasts support strategic decisions including optimal planting windows to avoid predicted stress periods and variety selection.

In-season Management (Real-time Severity): Daily stress severity measurements enable tactical adjustments such as irrigation scheduling, bio-stimulant applications and protective treatments. Real-time data helps growers respond quickly to developing stress conditions.

Post-season Analysis (Historical Characterization): Seasonal stress summaries provide comprehensive environmental characterization for yield analysis, insurance claims, variety performance evaluation, and planning improvements for subsequent seasons.

How to interpret insights

The abiotic stress model provides two insights for each crop stress type:

Daily stress severity index

• What it measures: current and near-term stress intensity using actual weather data (up to today) and 5-day forecasted conditions.
• Scale: 0 (no stress) to 9 (extreme severity)
• Primary use: In-season decision making
  • Generate real-time alerts
  • Adjust management practices during the growing season
  • Summarize seasonal environmental conditions post-harvest

Aggregated stress risk percentage

• What it measures: probability of stress events occurring at a specific week based on frequency of stress events over the historical period
• Scale: 0% to 100%
  • 0% = Stress threshold never exceeded at this timing in the historical period
  • 100% = Stress threshold exceeded every year at this timing in the historical period
• Primary use: Pre-season planning with the key goal of preventing vulnerable crop growth stages from coinciding with periods of elevated stress probability.
  • Optimize planting dates
  • Select appropriate genotypes/varieties
  • Schedule critical management practices to avoid high-risk periods

Limitations

• The indices are calculated using gridded daily weather data (10 x 10km), which could be subject to inaccuracies for some locations / fields.
• The stress indices may not accurately capture micro-climate variations or field-level and sub-field conditions that could influence crop stress levels.
• There could be genotypes that have different temperature thresholds and respond differently to stress, currently, the stress thresholds are not detailed at the genotype / variety level.
• The stress indices provide an estimate of the stress conditions based on weather data, but they do not directly measure actual crop growth or yield.

Last update: 06/10/2025