Effects of meteorological and ecological disturbances on tropical vegetation phenology
Abstract
Vegetation phenology is very sensitive and responsive to ecosystem drivers, short-term disturbances, meteorological extremes, and climate changes over time. It is a useful proxy for vegetation health, decline, and recovery. Satellite imagery can be used to study vast and remote tropical regions, though persistent cloud, smoke, and sensor error pose challenges that can limit data breadth. Common, long-term products such as MODIS and Landsat suffer from low spatial and/or temporal resolution and availability. This makes it difficult to capture subtle changes in heterogeneous tropical plant communities. We generated time series of imagery from ESAs Sentinel-2 MultiSpectral Instrument, which provides high spatial and temporal resolution. We computed the Normalized Difference Red Edge (NDRE) Index to construct phenological time series of tropical vegetation. We focused our analysis over select regions in Central and South America with varying topography, latitude, species distribution, land cover, and area. We sought to understand (1) dominant phenological patterns across space-time; (2) phenological responses to annual variability in meteorological conditions; and (3) climate conditions that drive anomalies and extremes in phenology. Using Google Earth Engine, we developed a Python-based computational workflow to analyze data from 20172020 at 20-meter spatial and 15-day temporal resolutions. Using ERA5 meteorological time series, we analyzed climate drivers of phenology. Significant NDRE anomalies are highly correlated with meteorological anomalies. While temperature often affects phenology immediately, precipitation has a lag effect. Precipitation and soil moisture conditions of past months are often the dominant drivers of negative anomalies in phenology, especially in dry tropical regions. Spatially variable phenology drivers and vegetation-specific responses can be used to improve our mechanistic understanding of phenological processes in tropical vegetation.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2021
- Bibcode:
- 2021AGUFM.B33D..04S