Procena snabdevenosti prirodnih travnjaka vodom primenom vremenske serije satelitskih snimaka

Abstract

INTRODUCTION and OBJECTIVES: Remote sensing is the process of detecting and monitoring the physical characteristics of an area by measuring its reflected and emitted radiation at a distance, and in a narrower sense includes the analysis and interpretation of various images of parts of the Earth's surface. The application of satellite images with modern technology and software is possible in all phases of research of various natural phenomena, and their analysis are carried out by computer-aided and visual procedures. The aim of this research is to present the basic aspects of remote sensing and modern technologies in assessing the water supply of natural grasslands, as well as the presentation of possibility for faster data collection while reducing costs, and easier understanding of the whole research area. MATERIAL and METHOD: The trial was set on March 22nd of 2022 on a natural grassland in the village of Mitrovo polje on the mountain Goč (43° 30′ 22′ N latitude, 20° 52′ 26′ E longitude, 700 m a.s.l.) in the Aleksandrovac municipality. Disturbed and undisturbed soil samples were collected to determine soil texture, soil water retention characteristics and soil chemical properties. Continuous measurement of soil moisture was performed with TDR probes in three replicates and sensors for monitoring soil temperature were installed. The satellite images (SENTINEL 2) with a resolution 10 m, in a time interval of about a week, created in relatively clear weather (cloudiness <30%), starting from March 22 to May 21st were used, while soil moisture data were collected on a daily basis. Data on precipitation and air temperature for the observed period were taken from the Agrosens portal. The supply of natural grassland with water and assessment of soil moisture are determined by the index of normalized difference vegetation index (NDVI), as the most applicable vegetation index. NDVI is defined as the ratio of differences between individual values of reflective wavelengths of near red and red radiation spectrum and their sum. Furthermore, for the analysis are also used MSAVI2 index, which is mainly used to analyse plant growth, estimate grass yield, monitor drought and soil erosion, and the optimized soil-adapted vegetation index (OSAVI), which is more sensitive to vegetation. The analysis of remotely sensed images of the investigated area and statistical analysis were conducted using QGIS tools. RESULTS and CONCLUSIONS: The results of the research show that from the beginning of grassland growth, the soil moisture ranged from 39% to 57%, which indicates that the lawn was well supplied with water. Water consumption was low at initial stage of grass growth. Soil water content at that period was high due to subsurface inflow coming from the higher parts of the terrain. The values of the NDVI index varied from 0.16 to 0.47, while the values for the MSAVI2 and OSAVI index ranged from 0.59-0.73 and 0.24-0.63, respectively. Lower index values at the beginning of the observed period (<0.16, <0.32, <0.30 for NDVI, MSAVI2 OSAVI, respectively) indicate a lack of green biomass. As the growth of natural grass increased, so do the values of all indices. Nevertheless, the obtained results show the changes of vegetation indices with the change of soil moisture, i.e. the values of the index increase due to the increase of soil moisture after precipitation and vice versa, which leads us to the conclusion that the application of remote sensing indices can be successfull for the estimate of different vegetation conditions, detection ofsoil moisture and biomass assessment.