Title : Managing the conditions of the processes of vegetation of grain crops using NDVI
Abstract:
Yield forecasting is an important tool for managing the production process and one of the mandatory elements of the precision farming system. Modeling of the effect of the current conditions on the vegetation of plants is realized by factor analysis of temperature, soil, climatic and other influences. An evaluation indicator should be selected as an independent variable of the mathematical model that most fully reflects the impact of the current conditions on the growth and development of culture, including the possibility of determining it in a remote format. As a criterion for optimizing the predictive mathematical model, the article considers the possibility of using deviations of the current NDVI values from the average long-term indicators. On the example of winter wheat and spring barley, the procedure for the formation of NDVI time series based on archival data from 2016-2020 is given. The approximation of long-term average data by an asymmetric gaussian is performed and the adequacy of mathematical models is checked, allowing
1. The task of managing the vegetation process can be implemented on the basis of predictive models obtained by factor analysis of the initial data.
2. A comparative analysis of the average long-term and current (vegetative season of 2021) NDVI indices for the studied crops, a diagram of NDVI anomalies (ΔNDVI) of the current vegetative season was obtained. The diagram is recommended for assessing the influence of external factors on the vegetative process.
3. The characteristic of ΔNDVI can be used as an independent variable (optimization criterion) in factor models for predicting the dynamics of the vegetation process.
