Model Computation Workflow

Step-by-step SIMPLE computation from input ingestion to final outputs

Workflow flowchart

1) Input ingestion: weather + management + species + cultivar + soil

2) Column mapping/cleaning and validation of dates and values

3) Daily ARID and water balance (Runoff, Drainage, TR, WAT, PAW, ET0)

4) Response factors: dTT, TT, fTemp, fHeat, fCO2, fWater

5) Canopy simulation: fSolar via I50A/I50B and stress effects

6) dBiomass and cumulative Biomass computation

7) Yield = Biomass × HI and run summary

8) Scenario batch runs and percent-change comparison to baseline

9) Uncertainty analysis (Monte Carlo) with quantiles/sensitivity

10) Scenario matrix full-factorial runs and ranking

Step details and formulas

Step 1: Data preparation

Required columns: DATE, TMAX, TMIN, RAIN, SRAD.

Optional columns: CO2, IRRIGATION, ET0.

Dates are converted to internal `IDATE` and sorted.

Formula: Tmean = (TMAX + TMIN) / 2

Step 2: Water balance and ARID

If ET0 exists in uploaded data, it is used directly; otherwise ET0 is computed by Priestley-Taylor.

Runoff is computed with Curve Number and deep drainage with DDC.

Daily ARID is derived from actual transpiration relative to ET0.

Formula: ARID = 1 - TR / ET0 ; fWater = max(0, 1 - S_Water * ARID)

Step 3: Thermal time and stress responses

Daily dTT and TT are computed for phenology.

Temperature response increases linearly up to Topt.

Heat response declines between MaxT and ExtremeT.

CO2 response is relative to 350 ppm and saturates near 700 ppm.

Formula: dTT = max(Tmean - Tbase, 0) ; TT(t) = Σ dTT

Step 4: fSolar canopy dynamics

fSolar uses two logistic curves for canopy build-up and senescence.

I50B is increased by water and heat stress and can accelerate decline.

When fSolar falls below the threshold, senescence maturity is triggered.

Formula: fSolar = min(fSolar1, fSolar2) × min(fSolar_water_prev, 1)

Step 5: Biomass growth and yield

Daily biomass growth uses RUE, radiation, and all response factors.

Cumulative biomass includes the initial biomass state.

Final yield is biomass multiplied by harvest index.

Formula: dBiomass = 10 × RUE × fSolar × SRAD × fCO2 × fTemp × min(fWater, fHeat)

Step 6: Scenario analysis

Each scenario mutates selected weather/parameter inputs and reruns the model.

Yield and biomass are compared to baseline in percent terms.

Formula: ΔYield% = (Yield_scenario / Yield_baseline - 1) × 100

Step 7: Uncertainty and sensitivity

Monte Carlo samples are drawn for RUE, S_Water, HI, Tsum, and CO2 ranges.

Key quantiles (5/25/50/75/95) and correlation-based sensitivity are computed.

Formula: Sensitivity(x) = corr(x, Yield) ; Quantiles = Qp(Yield)

Step 8: Scenario matrix

All combinations of temperature, CO2, rainfall, irrigation, RUE, and HI levels are generated.

Each combination is run independently and ranked in the final decision table.

Formula: N = nTemp × nCO2 × nRain × nIrr × nRUE × nHI