Head of Division Prof. Ratmir Poluektov

The main task of Division staff is development of the methodology of systems analysis directed to modeling of agroecosystems, and its application to decision support in agriculture. All the results were obtained in close connection with the Laboratory of Agroecosystem Simulation of Agrophysical Research Institute. Agroecosystems play an essential role in terrestrial water, energy and nutrient chains. They are the main source of organic matter for a majority of living creatures on the earth. On the other hand, an intensive exploitation of the agroecosystems leads ro many negative consequences to world population. Moreover, in the last decades a gap is turned out between the growing human pressure on agroecosystems and the forecast of both short- and long-time consequences of that pressure. So, the problem to maintain agriculture sustainability becomes one of the important tasks in the field of agroecosystem simulation.

Theoretical approach to modeling problem was formed in the last decade. All the models were divided into two classes. First of them, so-called "basic models" have an hour time step and contain a very detailed description of main processes in the system "soil-plant-atmosphere". They can be used as a tool for scientific investigation of the totality of the processes taking place in soils, plant canopies or pre-soil air. The models of the second class (applied models) have a daily time step and are based on a balanced description of the processes of energy- and mass-transfer in an agroecosystem. The main task of such models is their application to elaboration of control strategies in sustainable agriculture.

A basic model of plant production process and a family of applied models were developed. The models consist of some units; each of them describes a set of physically uniform processes. The main of them are: absorption of solar radiation by plant canopy; plant photosynthesis and respiration: water and heat dynamics in soil, plant and atmosphere; plant growth and development; nitrogen transformation and transfer in soil and plant; leaching of nitrates out of the rooting zone. Models are based on a new description of the following processes: (i) method of calculation of real values plant transpiration and soil evaporation is developed. It represents the modification and further development of well-known Penman-Monteith method, (ii) a concept of universal water stress index is proposed which takes into account both water deficit and overmastering (iii) and finally, adaptive distribution keys which were used to describe dry matter allocation in shoot and root.

As a result, a family of models is developed and implemented on a personal computer. The models are identified to the following cultures in several Russia's regions and abroad: winter and spring wheat, barley, maize, alfalfa and tomato. Note, that existing agroecological models are being developed in application to concrete site. At the same time a totality of processes taking place in a system "soil-plants-atmosphere" is closely connected with environmental conditions that vary from one location to another. So the problem arises:
"Is the model developed and calibrated with data from one specific site conditions applicable to the other geographical and climatic regions?". The answer to the question is that the model, which pretends to application to regional (or global) conditions, has to satisfy a set of quite definite criteria. The main of them are as follows. Model reactions to environment conditions must remain adequate only after change in physical site parameters (soil and weather) without model re-adaptation. It is proved that model family developed by us satisfies these criteria. The task of the group for the nearest future is to run the models in the framework of GIS.

GIS permits to present simulation results in geographical view. Another GIS feature is also important. It is the presentation of model results in time. Model permits to obtain not only static picture, but also the dynamics of the main processes taking place in an agroecosystem. It especially concerns the forecast of agroecosystem development. Our models can solve the following dynamic tasks:

• calculation of soil water dynamics and water availability;
  
• estimation of the dates of plant development;
  
• estimation of universal water stress index;
  
• forecast of the date of plant development during the vegetation;
  
• yield forecast (starting from the date of flowering).
  

All the results of the solution of these tasks can be presented in GIS as animation. In this case the calculations must be done in time-and-space format for the appropriate territory with various time steps.

(back to local indices)