| The rich, vast Amazonian ecosystem is directly and indirectly threatened by human activities; remote sensing serves as an essential tool for monitoring, understanding and mitigating these threats. A multi-faceted body of work is described here, addressing three major issues that employ and advance remote sensing techniques for the study of Amazônia and other tropical rainforest regions. In Chapter 2, canopy reflectance modeling and satellite observations were used to quantify the effect of epiphylls on remote sensing of humid forests. Modeling simulations demonstrated sensitivity of canopy-level near infrared and green reflectance to epiphylls on leaves. Time series of Moderate Resolution Imaging Spectrometer (MODIS) data corroborated the modeling results, suggesting a degree of coupling between epiphyll cover and vegetation indices which must be accounted for when using optical remote sensing in humid forests. In Chapter 4, 11 years (2000–2010) of MODIS land surface temperature (LST) data covering the entire Amazon basin were used to ascertain the role of heat stress during droughts in 2005 and 2010. Preliminary accuracy assessments showed that LST data provided reasonably accurate estimates of daytime air temperatures (RMSE = 1.45°C; Chapter 3). There were moderate to strong correlations between LST-based air temperature estimates and tower measurements (mean r = 0.64), illustrating a sensitivity to temporal variability. During both droughts, MODIS LST data detected anomalously high daytime and nighttime canopy temperatures throughout drought-affected regions. Multivariate linear models of LST and precipitation anomalies explained 65.1% of the variability in forest biomass losses, as determined from a wide network of forest inventory plots. These results suggest that models should incorporate both heat and moisture to predict drought effects on tropical forests. In Chapter 5, I performed high spatial and temporal resolution modeling of carbon stocks and fluxes in the state of Rondônia, Brazil for the period 1985–2009. Based on this analysis, Rondônia contributed ∼4% of pan-tropical humid forest deforestation emissions while carbon uptake by secondary forest was negligible due to limited spatial extent and high turnover rates. Spatial analysis of land cover change demonstrated the necessity for fine resolution carbon monitoring in tropical regions dominated by non-mechanized, smallholder land uses. |
