Estimation of Sunflower Crop Production Based on Remote Sensing Techniques

Abstract

The study used the remote sensing method (Sentinel 2) to analyze the sunflower crop and to estimate the production. The study area was within the DES, ULS 'King Mihai I' from Timisoara, Romania. Eight series of images were taken (April 06 - August 07, 2022). Based on the spectral information, the NDMI, NDVI, NPCRI and NBR indexes were calculated. Spline models best described the variation of index values in relation to time (t, days) during the study period, ε = −0.04286 for NDMI, ε = 0.01172 for NDVI, ε = 0.00537 for NPCRI, respectively ε = −0.08481 for NBR. Very strong correlations were found between NDVI and NDMI (r=0.975), between NBR and NDMI (r=0.997), and between NBR and NDVI (r=0.967), p<0.001. Strong correlation was recorded between NDVI and NPCRI (r=-0.881), p<0.01. Moderate correlations were found between NDMI and t (r=0.729), between NBR and t (r=0.752), between NPCRI and NDMI (r=-0.776), and between NBR and NPCRI (r=-0.762), p<0.05. The regression analysis facilitated the estimation of the production based on calculated indices, under conditions of statistical safety.

Normalised difference moisture index in water stress assessment of maize crops

“…There are some interesting and valuable scientific studies related to the implementation of NDMI in agricultural science. The study by Herbei et al (2023) provided novel insights on the correlation between different spatial vegetation indices, as well as on the possibility of their implementation for sunflower yield prediction. The researchers found very strong correlation between NDMI and NDVI (correlation coefficient 0.975), and NDMI and NBR (correlation coefficient 0.997), which are normalised difference vegetation index (Huang et al, 2021) and normalised burn ratio, respectively (Delcourt et al, 2021).…”