The Influences of Meteorological Factors on Mapping Forest Stock Volume With Sentinel-1 Images

Synthetic Aperture Radar (SAR) has a distinct advantage over optical remote sensing due to its ability to map forest s tock volume (FSV) relatively accurately in complex forests, while being less affected by weather conditions. However, significant discrepancies persist in the estimated FSV when comparing results derived from multiple dual-polarization SAR images acquired within a short time frame. This inconsistency severely limits the reliability and validity of FSV mapping. To investigate the influence of meteorological factors on FSV mapping accuracy, we collected 30 dual-polarization SAR images (Sentinel-1) throughout the year and simultaneously derived meteorological parameters (TMP, G ST, GST005, SM005, SHU and WIN) from real-time datasets provided by China Meteorological Administration Land Data Assimilation System. To quantitatively assess the interference caused by meteorological factors on features and FSV mapping accuracy, we propose a novel metric called the meteorological disturbance in dex (MDI). The results revealed a strong correlation (ranging from 0.56 to 0.65) between four meteorological factors (GST, GST00 5, TMP, and SHU) and the backscattering coefficient. Among the 30 images analyzed in this study, the coefficient of determination (R2) for mapping FSV ranged between 0.03 and 0.34 due to variations in feature sensitivity. Additionally, temperature-related meteorological factors were found to positively influence feature sensitivity with lower temperatures amplifying this interference effect, whereas humidity-related factors exhibited contrasting effects on feature sensitivity levels. Furthermore, the application of MDI can help identify stable features during periods without available imagery, thereby offering significant potential for improving the accuracy and reliability of FSV mapping using SAR images.