Rainfall-caused water film on canopy surface biases remotely-sensed vegetation greenness

Abstract

Remotely-sensed vegetation greenness exhibits obvious differences before and after the short-term heavy rainfall event. This short communication reports that residual water film on the canopy caused by precipitation directly affects the spectral signal of vegetation, which in turn biases observed vegetation greenness. We combined ground measurements, unmanned aerial vehicle (UAV) measurements, three-dimensional radiative transfer model (RTM) simulations, and satellite observations to assess this phenomenon and investigate the impact of water film on vegetation indices (VIs) across different scales and vegetation types. Our findings demonstrate that precipitation exerts a quick and significant influence on the spectral characteristics of canopy components. The presence of water can lead to reflectance attenuation in soil and vegetation components across the entire visible and infrared spectrum, particularly in the near-infrared (NIR) band, with reductions of up to 0.12. The reduction in VI measures after precipitation can be explained by an imbalance in the magnitude of visible and NIR reflectance attenuation caused by the canopy water film. Taking the difference vegetation index (DVI) as an example, the decrease in NIR reflectance is more pronounced than that of red reflectance in the presence of water, resulting in DVI decreasing after a rainfall event. Satellite observations indicated that NDVI and EVI could decrease even exceeding 0.3 in some pixels after short-term rainfall. This work reveals the water film as a possible factor contributing to the bias between true and observed vegetation greenness. The spatial distribution of rainfall seasonality exhibits variation across different regions worldwide, and mitigating the impact of water film on vegetation greenness is essential for enhancing the precision and reliability of global vegetation dynamics monitoring.