Seasonal variations in vegetation water content retrieved from microwave remote sensing over Amazon intact forests - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue Remote Sensing of Environment Année : 2023

Seasonal variations in vegetation water content retrieved from microwave remote sensing over Amazon intact forests

Wigneron J.-P.
Frédéric Frappart
  • Fonction : Auteur
  • PersonId : 1338014
  • IdHAL : ffrappart

Résumé

Vegetation optical depth (VOD) is seasonally sensitive to plant water content and aboveground biomass. This index has a strong penetrability within the vegetation canopy and is less impacted by atmosphere aerosol contamination effects, clouds and sun illumination than optical vegetation indices. VOD is thus increasingly applied in ecological applications, e.g., carbon stock, phenology and vegetation monitoring. However, VOD retrieval over dense forests is subject to uncertainties caused by the thick canopy and complex multiple scattering effects. Thus, a comprehensive evaluation of VOD products over dense forests is needed for effective and accurate applications. This study evaluated the seasonal variations of eight recently developed/reprocessed VOD products at different frequencies (e.g., Ku-, X-, C- and L-band) over Amazon intact forests, supported by the ORCHIDEE-CAN-NHA model-simulated vegetation water content. Furthermore, we also explored the potential causes of VOD retrieval issues, in terms of retrieval algorithm uncertainties. We first confirmed that soil water availability dominated seasonal dynamics of vegetation water content over Amazon intact forests. This was verified by model-simulated vegetation water content and by C-band radar backscatter observations. Generally, evening or midday vegetation water content shows higher correlations with soil moisture than morning or midnight vegetation water content. In terms of ability of morning or midnight VOD products to follow the seasonality of soil moisture, active microwave ASCAT-IB C-VOD (median seasonal correlation with soil moisture (R) ∼ 0.50) outperforms the passive microwave VOD products, followed by passive microwave AMSR2 X-VOD (R ∼ 0.26) and VODCA X-VOD (R ∼ 0.16). However, SMOS-IC L-VOD (R ∼ −0.15) and AMSR2 C1-VOD (R ∼ −0.20) show obviously negative seasonal correlations with soil moisture across most pixels. This implausible behavior is likely to be caused by the inappropriate setting of time-invariant scattering effects in the passive microwave VOD retrieval algorithms, which could lead to an overestimation of the VOD amplitude during dry seasons. Thus, we recommend that the seasonal scattering effects be considered in the passive microwave VOD retrieval algorithms. These findings can contribute to the improvement of VOD retrieval algorithms and help with the development of their ecological applications over Amazon dense forests.

Dates et versions

hal-04095055 , version 1 (11-05-2023)

Identifiants

Citer

Huan Wang, Wigneron J.-P., Philippe Ciais, Yitong Yao, Lei Fan, et al.. Seasonal variations in vegetation water content retrieved from microwave remote sensing over Amazon intact forests. Remote Sensing of Environment, 2023, 285, pp.113409. ⟨10.1016/j.rse.2022.113409⟩. ⟨hal-04095055⟩
49 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More