About possibilities of apple trees flowering date detection based on modis data

Operational monitoring of orchards is the basis for decision-making in management and correction of agricultural practices during fruits production. Especially important is the monitoring of phenological development of trees, which predefines agrotechnological operations, and fertilizers application dozes, and dates. The monitoring results as well can be used for forecasting fruit yield and planning fruit production and fertilizer needs. At present, operational monitoring of large orchards is practically not conducted due to its high cost and time. The proposed article deals with the analysis of possibilities of using weekly MODIS composites (NDVI and reflection in separate bands of the sensor) for operational monitoring of apple flowering date. The main source of information was the data from the Internet-service "VEGA" as well as the data of field surveys at the key sites in the Krasnodar region of Russia for the period from 2001 to 2019. The conducted studies showed that the dates of apple flowering determined in the field conditions are poorly correlated with both NDVI values and reflection of electromagnetic waves in individual MODIS survey bands. The best indicator of apple flowering dates is the reflection value in the MODIS NIR survey band. Based on this value, the date of apple flowering in the study area can be predicted with an error of 2 weeks. The large error is related both to the specifics of the used satellite data (weekly compositions) and to the specifics of the apple orchard as a monitoring object. When organizing satellite-based monitoring of apple flowering dates, it is necessary to take into account the age of the trees, peculiarities of the apple tree growing techniques, as well as the condition of the ground cover between the rows of the garden (presence of herbaceous vegetation there and frequency of its mowing). The accuracy of detection can theoretically be increased by using daily satellite images. However, the interfering influence of cloud cover does not allow using such data in the research region. © 2020 International Multidisciplinary Scientific Geoconference. All rights reserved.