Influence of surface water variations on VOD and biomass estimates from passive microwave sensors

Emma Bousquet,Yann Kerr,Arnaud Mialon,Fabien Wagner,Catherine Prigent,Nemesio Rodriguez-Fernandez

doi:10.5194/egusphere-egu21-5786

Abstract

&lt;p&gt;Vegetation optical depth (VOD) is a remotely sensed indicator characterizing the attenuation of the Earth's thermal emission at microwave wavelengths by the vegetation layer. At L-band, VOD can be used to estimate and monitor aboveground biomass (AGB), a key component of the Earth's surface and of the carbon cycle. We observed a strong anti-correlation between SMOS (Soil Moisture and Ocean Salinity) L-band VOD (L-VOD) and soil moisture (SM) anomalies over seasonally inundated areas, confirming previous observations of an unexpected decline in K-band VOD during flooding (Jones et al., 2011). These results could be, at least partially, due to artefacts affecting the retrieval and could lead to uncertainties on the derived L-VOD during flooding. To study the behaviour of SMOS satellite L-VOD retrieval algorithm over seasonally inundated areas, the passive microwave L-MEB (L-band Microwave Emission of the Biosphere) model was used to simulate the signal emitted by a mixed scene composed of soil and standing water. The retrieval over this inundated area shows an overestimation of SM and an underestimation of L-VOD. This underestimation increases non-linearly with the surface water fraction. The phenomenon is more pronounced over grasslands than over forests. The retrieved L-VOD is typically underestimated by ~10% over flooded forests and up to 100% over flooded grasslands. This is mainly due to the fact that i) low vegetation is mostly submerged under water and becomes invisible to the sensor; and ii) more standing water is seen by the sensor. Such effects can distort the analysis of aboveground biomass (AGB) and aboveground carbon (AGC) estimates and dynamics based on L-VOD. Using the L-VOD/AGB relationship from Rodriguez-Fernandez et al. (2018), we evaluated that AGB can be underestimated by 15/20&lt;sup&gt;&lt;/sup&gt;Mg ha&lt;sup&gt;-1&lt;/sup&gt; in the largest wetlands, and up to higher values during exceptional meteorological years. Such values are more significant over herbaceous wetlands, where AGB is ~30 Mg ha&lt;sup&gt;-1&lt;/sup&gt;, than over flooded forests, which have typical AGB values of 150-300 Mg ha&lt;sup&gt;-1&lt;/sup&gt;. Consequently, to better estimate the global biomass, surface water seasonality has to be taken into account in passive microwave retrieval algorithms.&lt;/p&gt;

Highlights

Large-scale monitoring of vegetation cover is crucial for under standing its behaviour and its links with climate evolution, extreme events, and land cover changes (Piao et al, 2019; Qin et al, 2019)
We highlighted the anomalous decrease of Vegetation optical depth (VOD) during flooding; and we showed with a modelling experiment that this phe nomenon was linked to the influence of standing water temporal vari ations
SMOS Level2, Level3, and SMAP operational algorithms take the major water bodies into account with a static map, but this study showed the importance of considering the temporal dynamics of water extent

Summary

Introduction

Large-scale monitoring of vegetation cover is crucial for under standing its behaviour and its links with climate evolution, extreme events, and land cover changes (Piao et al, 2019; Qin et al, 2019). Visible frequencies have predominantly been used for these applica tions, thanks to the high spatial resolution of optical instruments. They are impaired by their inability to penetrate clouds and dense vegetation. The low frequency L-band (1.4 GHz) VOD (L-VOD) measured with SMOS satellite was proven to be highly sensitive to aboveground biomass (AGB) in Africa, with less saturation over dense forests than optical indices and than C- or X-VOD (Rodríguez-Fernandez et al, 2018). A strong correlation was found at the global scale between L-VOD and two AGB datasets (R = 0.91–0.94), but was shown to be highly dependent on the vegetation type

Objectives

Methods

Results

Discussion

Conclusion