VEGETATION TEMPORAL VARIATION ANALYSIS WITHIN "PADUREA CENAD" PROTECTED AREA BASED ON REMOTE SENSING

Abstract

The study used the technique based on remote sensing to analyze the temporal variability within the "Cenad Forest" Protected Area, Timis County, Romania. The Sentinel 2 satellite system was used to obtain satellite images. Images were taken in the four seasons, the period 2020-2022. The result was a package of 12 sets of satellite data, one for each anotype and year of the study period. Based on the spectral values, the NDMI, NDVI and NBR indices were calculated (series of 908 data for each index, season and year). Descriptive statistical analysis was used for the primary evaluation of the data. In the case of 2020, the variation of the NDVI index in relation to NDMI and of the NBR index in relation to NDMI and NDVI was described by polynomial equations of degree 2, with the exception of moment t3, when NBR in relation to NDVI was described by a linear equation. In the case of the year 2021, the variation of the NBR index in relation to the NDVI (at all four moments of determination) was described by linear equations, and in the case of the other indices the variation was described by polynomial equations of the 2nd degree. In the case of 2022, the variation of the NBR-t1 index in relation to NDVI-t1 was described by a polynomial equation of the 2nd degree, and in all other cases, the variation of the indices was described by linear equations. In the case of 2022, the average values of the indices showed lower levels than the average values recorded in 2020 and 2021. The comparative analysis of the data series, for each index, from the years 2020 and 2021 compared to the year 2022, led to statistically safe differences (p<0.001). An exception was registered in the case of NDMI-t4 (year 2020) and NDVI-t1 (2021) when the differences did not present safety (p>0.05).