Search

The capacity of orbital sensors to monitor vegetation functioning and phenology in the Amazon rainforest using vegetation indices (VIs) has been broadly challenged in recent years. In particular, recent studies indicate that artifacts associated with sun-sensor geometry are likely to have major influence on the variability of some VIs. Nevertheless, the magnitude of this influence and the impacts on different VIs are still poorly understood. This study evaluates the scaling and magnitude of the phase angle variation effects on six different VIs, as well as in the individual spectral bands used for computing these indices. Our results show a significant and consistent relationship between phase angle and VI anomalies. The scaling (i.e. the change rate of VI anomaly with phase angle) of this relationship varies according to spectral bands and VIs. Median anomalies in individual Moderate Resolution Imaging Spectroradiometer (MODIS) bands showed a variation of up to −0.045 degree−1, while median VI anomaly variations reached −0.047 degree−1. The scaling of all relationships was shown to be constant throughout the year. Average phase angle values in the Amazon basin ranged from approximately 25 degrees in October to 41 degrees in June. Such variation can, on average, cause changes of up to 0.75 in some VIs, therefore having major impacts on the interpretation of the relationships between VIs and forest phenology.

During recent years, many studies have been undertaken to investigate how spectral characteristics of forests can provide information on spatial patterns of tree species diversity (TSD). Important advances have been made, and significant relationships between TSD and remotely sensed indicators of net primary productivity and environmental heterogeneity have been reported. However, the season-dependence of these relationships has not yet been fully investigated, and the influence of phenology remains poorly understood. In this study, we aim to assess how the relationships between remote sensing indicators and TSD depend on the season of the year. TSD measures, including species richness, Shannon’s diversity and Simpson’s diversity, were determined for 82 field plots in the Afromontane cloud forests of Taita Hills, Kenya. A time series of 15 Landsat images were used to calculate a set of spectral and heterogeneity metrics. The relationship between remote-sensing metrics and TSD measures was analysed by simple and multivariate regression analysis. We conclude that the relationships between remote-sensing metrics and TSD are season-dependent. Hence, it is demonstrated the date of image acquisition is an important aspect to be considered in biodiversity studies. Given that the dependence of the relationships is closely linked to climate seasonality defining vegetation phenology, the relationships may also vary according to geographical conditions.

The capacity of orbital sensors to monitor vegetation functioning and phenology in the Amazon rainforest using vegetation indices (VIs) has been broadly challenged in recent years. In particular, recent studies indicate that artifacts associated with sun-sensor geometry are likely to have major influence on the variability of some VIs. Nevertheless, the magnitude of this influence and the impacts on different VIs are still poorly understood. This study evaluates the scaling and magnitude of the phase angle variation effects on six different VIs, as well as in the individual spectral bands used for computing these indices. Our results show a significant and consistent relationship between phase angle and VI anomalies. The scaling (i.e. the change rate of VI anomaly with phase angle) of this relationship varies according to spectral bands and VIs. Median anomalies in individual Moderate Resolution Imaging Spectroradiometer (MODIS) bands showed a variation of up to −0.045 degree−1, while median VI anomaly variations reached −0.047 degree−1. The scaling of all relationships was shown to be constant throughout the year. Average phase angle values in the Amazon basin ranged from approximately 25 degrees in October to 41 degrees in June. Such variation can, on average, cause changes of up to 0.75 in some VIs, therefore having major impacts on the interpretation of the relationships between VIs and forest phenology.

During recent years, many studies have been undertaken to investigate how spectral characteristics of forests can provide information on spatial patterns of tree species diversity (TSD). Important advances have been made, and significant relationships between TSD and remotely sensed indicators of net primary productivity and environmental heterogeneity have been reported. However, the season-dependence of these relationships has not yet been fully investigated, and the influence of phenology remains poorly understood. In this study, we aim to assess how the relationships between remote sensing indicators and TSD depend on the season of the year. TSD measures, including species richness, Shannon’s diversity and Simpson’s diversity, were determined for 82 field plots in the Afromontane cloud forests of Taita Hills, Kenya. A time series of 15 Landsat images were used to calculate a set of spectral and heterogeneity metrics. The relationship between remote-sensing metrics and TSD measures was analysed by simple and multivariate regression analysis. We conclude that the relationships between remote-sensing metrics and TSD are season-dependent. Hence, it is demonstrated the date of image acquisition is an important aspect to be considered in biodiversity studies. Given that the dependence of the relationships is closely linked to climate seasonality defining vegetation phenology, the relationships may also vary according to geographical conditions.

The capacity of orbital sensors to monitor vegetation functioning and phenology in the Amazon rainforest using vegetation indices (VIs) has been broadly challenged in recent years. In particular, recent studies indicate that artifacts associated with sun-sensor geometry are likely to have major influence on the variability of some VIs. Nevertheless, the magnitude of this influence and the impacts on different VIs are still poorly understood. This study evaluates the scaling and magnitude of the phase angle variation effects on six different VIs, as well as in the individual spectral bands used for computing these indices. Our results show a significant and consistent relationship between phase angle and VI anomalies. The scaling (i.e. the change rate of VI anomaly with phase angle) of this relationship varies according to spectral bands and VIs. Median anomalies in individual Moderate Resolution Imaging Spectroradiometer (MODIS) bands showed a variation of up to −0.045 degree−1, while median VI anomaly variations reached −0.047 degree−1. The scaling of all relationships was shown to be constant throughout the year. Average phase angle values in the Amazon basin ranged from approximately 25 degrees in October to 41 degrees in June. Such variation can, on average, cause changes of up to 0.75 in some VIs, therefore having major impacts on the interpretation of the relationships between VIs and forest phenology.

The capacity of orbital sensors to monitor vegetation functioning and phenology in the Amazon rainforest using vegetation indices (VIs) has been broadly challenged in recent years. In particular, recent studies indicate that artifacts associated with sun-sensor geometry are likely to have major influence on the variability of some VIs. Nevertheless, the magnitude of this influence and the impacts on different VIs are still poorly understood. This study evaluates the scaling and magnitude of the phase angle variation effects on six different VIs, as well as in the individual spectral bands used for computing these indices. Our results show a significant and consistent relationship between phase angle and VI anomalies. The scaling (i.e. the change rate of VI anomaly with phase angle) of this relationship varies according to spectral bands and VIs. Median anomalies in individual Moderate Resolution Imaging Spectroradiometer (MODIS) bands showed a variation of up to −0.045 degree−1, while median VI anomaly variations reached −0.047 degree−1. The scaling of all relationships was shown to be constant throughout the year. Average phase angle values in the Amazon basin ranged from approximately 25 degrees in October to 41 degrees in June. Such variation can, on average, cause changes of up to 0.75 in some VIs, therefore having major impacts on the interpretation of the relationships between VIs and forest phenology.