Crop Growth Model

2020-09-30 16:34 irripro

The basis of fertigation decision-making is to obtain the result of crop growth state through mathematical model simulation. The model mainly simulates the effects of climatic, soil and management conditions of crops, simulates physiological processes such as assimilation, respiration, transpiration, dry matter distribution and nitrogen dissipation through four modules of crops, climate, soil and nutrients, simulates the growth period, biomass, yield and leaf area index of crops, and can simulate crop growth and yield formation at three levels of water   potential, water limitation and nutrient limitation. Based on potential yield or water limited yield, the nutrient limited yield is simulated by the stress effect of soil available nutrients on crop growth. The crop growth model is the key module in the intelligent fertigation decision engine, and the module optimization and algorithm supplement are mainly for this module.

The total biomass yield in crop growth period is equal to the average daily biomass yield multiplied by the total growth duration, so it is necessary to better simulate the growth duration for reliable biomass and yield prediction.

Sowing date or emergence time is an important input parameter of mathematical model. If sowing date is used as input, the emergence date is calculated according to the sum of temperatures from sowing to emergence TSUMEM. In this case, it is necessary to carry out field experiments on the same crop variety, and calibrate TSUMEM with the observed sowing and emergence dates.

According to two parameters describing the sum of the required effective accumulated temperatures, the crop growth period of the mathematical model is calculated. TSUM1 describes the accumulation of temperatures from emergence to flowering, and TSUM2 describes the accumulation   of temperatures from flowering to maturity. The daily effective temperature is equal to the daily average temperature minus the reference temperature. The DVS value of crop phenology represents a development stage, and DVS is equal to the ratio of the sum of accumulated effective temperatures to the parameters of TSUM1 and TSUM2. Here, DVS=1 at flowering stage and DVS=2 at mature stage are commonly used.

It should be noted that the growth period of crops in the engine can be divided into multiple growth periods according to different crops, but the sum of effective accumulated temperatures in each growth period needs to be realized by increasing the number of accumulated temperatures in stages.

Soil nutrient availability SAN in crop model is an important factor used to simulate crop growth, and the change of SAN can be easily converted into the output parameters of crop model. Using remote sensing (UAV) data to solve this problem. Data assimilation method combining crop model with time series remote sensing observation is helpful to extrapolate simulation from single point data application to large area application. By using EnKF filtering method, combining the mathematical model of crop growth with remote sensing observation, the simulation of SAN content change in the whole plough layer during pixel-scale growth was realized. The model is optimized and improved by using big data analysis, so as to establish a stable relationship between crop growth and SAN content.

Vegetation coverage index (NDVI) is used to estimate leaf area index, and NDVI is calculated and analyzed by overlapping NIR near infraRED band (760~900 nm) and red red band (630~690 nm). To find LAI from NDVI, using soil nutrient module, we can estimate nutrient-constrained yield based on potential or water-constrained yield. The main parameters of the original mathematical models of different crops were calibrated under different meteorological, soil and management conditions by assimilating the data collected in the field and remote sensing data.

Fertilization and simulated yield were selected as indicators to test whether soil nutrient stress would affect digital growth simulation based on crop growth. The new method is applied to variable fertilization, which improves the effect of accurate variable fertilization.