Crop Modelling Deserves More Attention from Authorities | 2018-02-10 | daily-sun.com

## Crop Modelling Deserves More Attention from Authorities

### 10 February, 2018 12:00 AM

Prior to grow a crop a farmer has to have a plan in his mind. The plan must be how to harvest the maximum output through the efficient utilisation of minimum input. To realise that plan he has to advance through a particular mental set up which could be designated as a kind of simple crop modelling. Or more precisely, this could be said that the farmer is growing crop through a mental modelling of his own. Since the beginning of agriculture, all the farmers are following the same kind of modelling either knowingly or unknowingly. In fact this kind of modelling is experience-based. But in a digital world, the farmers are waiting to have a tool that could be used to solve his agronomic and the related problems.

Anyway, the scientists are in progress in parallel with the development of the Computer Technology. There are already some significant developments of getting help in growing crops through the use of the digital technology. But they are not much popular among the targeted stakeholders. The most of the scientists are still using some conventional statistical tools in their agricultural research.  Though those tools are a kind of model but do not go beyond the boundary of the so-called “cause and effect” relationship. These models have a little ability to consider the details of the biological and physiological system of a crop in progress (growing) in the field. For example rice yield can be increased with the increased application Urea (N=Nitrogen) dose provided the variety have the affinity to N.  The urea dose is the “cause” and rice yield is the “effect”. The yield will increase up to a certain limit and then followed by a decreasing trend. So the cause and effect in this case  follows a mathematical relationship like Y = a + bx +cx2 where Y is yield, “x” is the increased urea (N) doses, “a” is intercept, “b” and “c” are the constants estimated from this quadratic relationship between yield and N doses.

We could easily estimate the yield against any value of N doses in this equation. This is a static model and it has got a lot limitations. The phenomenon of growth throughout the crop growth duration could not be explained to utilise for further development of cultural management practices relevant to the crop. More so, the result could not be extrapolated precisely beyond the periphery where the study is executed because the parameters a, b, and c would vary with the changing environments, soil types, cultural practices even with the crop variety itself. To validate the result of a specific static model in the larger extent numerous experiments under varying conditions are to be done to have the representative parameters which is comprehensively time consuming and costly too. Despite the limitations this kind of models are extensively used by the agricultural researchers.

Therefore, agricultural scientists were in a try to overcome the limitation seeking help from the mathematicians, engineers and computer experts for several decades. They consider the growing of crop as a system as the Engineers do in the case of their engineering projects/machineries. Thus the two complementary words “system and simulation” were borrowed from the field of Engineering to the field of crop production technology. The “System” simply represents the details of the reality of phenomenon as the sequence of growing rice crops to a rice scientist which depends on the environments etc.

The System includes photosynthesis, respiration, Nutrient and water uptake from the soil, growing conditions such as solar radiation, temperature, drought, salinity. It also includes partitioning of produced dry matter in the form of shoot weight root weight, the development phases like vegetative phase, reproductive stages, and grain weight at maturity etc. Those systems are framed in some mathematical traps (formulae and equations) in the systematic way. Then they are arranged in a computer programme following a relational diagram designed through the joint effort of a crop expert and a programmer. Crop and climate parameters are used as the input to get the output i.e. yield. If we want to know the variety-wise performance, then, obviously some of the variety-specific data should be used as input. This kind of model is quite complex and designated as dynamic or analytical model.

In a dynamic model, a change of a variable (for example crop biomass per unit area) for a brief unit of time (Δt) is considered as constant. The change of variable over a unit time is called rate variable. The time span may be from several hours to several days depending on the circumstances and the objective of the study. Now the question:consider “increment of biomass (variable)” in a dynamic crop model? Let us consider Wt is the biomass of a crop per unit area for the time “t”. For the brief time, no change of Wt is considered (static variable). But by differentiation (from calculus) the rate variable (dw/dt) over “t” time could be estimated. Then the increased weight would be Wt + t = Wt + Δt (dw/dt). Thus another rate variable will be estimated from this Wt + Δt. And the process will continue to reach the final weight until the end of growth duration.
The dynamic crop models initiated its journey through a simple water balance model as early as in the 1960s. The first attempt at modelling of the photosynthetic rates of crop canopy was done in mid-sixties by a famous Dutch Scientist de Wit.
The Dutch scientists like de Wit, Penning de Vries, Gourdian etc. had their significant involvement in the development of dynamic crop simulation modelling. They were fortunate that their works were in progress with the development of computer technology. To simulate the growth behaviour they used the Continuous Simulation Modelling Programme III (CSMPIII) developed by engineers for their purpose.  The original programme was then used to simulate plant or crop growth and was designated as Contiguous Simulation Modelling Programme for Plant. During the late 1980s, Penning de Vries and his group worked at IRRI with the scientists of the rice world. In fact, they were able to make a significant development in the field of crop modelling for rice. They are the one developed ORYZA ver. 3 evolved from its earlier versions (ORYZA2000 v2.13, v2.12, v2 and v1), which were developed from the earlier version ORYZA1, ORYZA_W and ORYZA_N. ORYZA2000 v2.12 and V2.13 were respectively integrated into cropping system models APSIM (The Agricultural Production System Simulator) and DSSAT (The Decision Support System for Agro-technology Transfer) as APSIM-ORYZA (2005) and DSSAT-ORYZA (2012) to handle rice simulations in different cropping systems (source internet). These models are now in use for research, teaching, agronomic management, crop forecasting and policy analysis.
Now, let us have a glimpse how this crop modelling study can be utilised in the field of agricultural research. There is a significant yield gap between the research and farmers’ field. The simulation modelling study can help to reduce the gap. The using of the parameters like planting density, fertiliser dose, time of fertiliser application, irrigation requirement, efficient cropping system, etc. in a proper crop modelling programme, a scientist could find the appropriate agronomic practices without conducting costly and extensive field studies. Even how a crop would perform in its new environment can be predicted. The model related to the impact of biotic and abiotic stresses could be incorporated with the main model to understand the fate of crop under stress environment.  Even the probability of pest resurgence could be predicted. Thus unnecessary pesticide application could be avoided. Some of the models are extensively in use to understand the global warming impact of crop yield. Now many scientists have an understanding how the crop yield would be reduced if no stress (drought, extreme temperature, flood etc.) tolerant varieties are available in near future.   Therefore, the scientists are getting ready to develop new stress tolerant crop varieties to cope with the increasing climate adversaries.
The unique thing is, most of the research works are done with the help of computers. Anyway, the initial crop parameter is collected through the studies under controlled or under field conditions. So a lot of manpower could be saved through the crop modelling studies.
Despite a lot of research, the application of crop modelling is still limited mostly within the researchers. Even scientists are yet to communicate the technology properly to the farmers except for few instances like fertiliser guidance for rice cultivation in some countries. This is not due to the weakness of the system.  The quality of output data depends on the quality of input data. It means the agricultural models are the reflections of the biological/physiological system which are yet to be explained properly. Anyway, the scientists are not disheartened. They are still working for an excellent matching between the crop phenomenon, mathematical interpretation and the computer programming so that it could be used with a little misunderstanding.