BackgroundUnderstanding the drivers of large-scale vegetation change is critical to managing landscapes and key to predicting how projected climate and land use changes will affect regional vegetation patterns. This study aimed to improve our understanding of the role, magnitude and spatial distribution of the key environmental factors driving vegetation change in southern African savanna, and how they vary across physiographic gradients.Methodology/Principal FindingsWe applied Dynamic Factor Analysis (DFA), a multivariate times series dimension reduction technique to ten years of monthly remote sensing data (MODIS-derived normalized difference vegetation index, NDVI) and a suite of environmental covariates: precipitation, mean and maximum temperature, soil moisture, relative humidity, fire and potential evapotranspiration. Monthly NDVI was described by cyclic seasonal variation with distinct spatiotemporal patterns in different physiographic regions. Results support existing work emphasizing the importance of precipitation, soil moisture and fire on NDVI, but also reveal overlooked effects of temperature and evapotranspiration, particularly in regions with higher mean annual precipitation. Critically, spatial distributions of the weights of environmental covariates point to a transition in the importance of precipitation and soil moisture (strongest in grass-dominated regions with precipitation<750 mm) to fire, potential evapotranspiration, and temperature (strongest in tree-dominated regions with precipitation>950 mm).Conclusions/SignificanceWe quantified the combined spatiotemporal effects of an available suite of environmental drivers on NDVI across a large and diverse savanna region. The analysis supports known drivers of savanna vegetation but also uncovers important roles of temperature and evapotranspiration. Results highlight the utility of applying the DFA approach to remote sensing products for regional analyses of landscape change in the context of global environmental change. With the dramatic increase in global change research, this methodology augurs well for further development and application of spatially explicit time series modeling to studies at the intersection of ecology and remote sensing.

Crop classification provides relevant information for crop management, food security assurance and agricultural policy design. The availability of Sentinel-1 image time series, with a very short revisit time and high spatial resolution, has great potential for crop classification in regions with pervasive cloud cover. Dense image time series enable the implementation of supervised crop classification schemes based on the comparison of the time series of the element to classify with the temporal signatures of the considered crops. The main objective of this study is to investigate the performance of a supervised crop classification approach based on crop temporal signatures obtained from Sentinel-1 time series in a challenging case study with a large number of crops and a high heterogeneity in terms of agro-climatic conditions and field sizes. The case study considered a large dataset on the Spanish province of Navarre in the framework of the verification of Common Agricultural Policy (CAP) subsidies. Navarre presents a large agro-climatic diversity with persistent cloud cover areas, and therefore, the technique was implemented both at the provincial and regional scale. In total, 14 crop classes were considered, including different winter crops, summer crops, permanent crops and fallow. Classification results varied depending on the set of input features considered, obtaining Overall Accuracies higher than 70% when the three (VH, VV and VH/VV) channels were used as the input. Crops exhibiting singularities in their temporal signatures were more easily identified, with barley, rice, corn and wheat achieving F1-scores above 75%. The size of fields severely affected classification performance, with ~14% better classification performance for larger fields (>1 ha) in comparison to smaller fields (<0.5 ha). Results improved when agro-climatic diversity was taken into account through regional stratification. It was observed that regions with a higher diversity of crop types, management techniques and a larger proportion of fallow fields obtained lower accuracies. The approach is simple and can be easily implemented operationally to aid CAP inspection procedures or for other purposes.

No abstract

Deconstructing the drivers of large-scale vegetation change is critical to predicting and managing projected climate and land use changes that will affect regional vegetation cover in degraded or threated ecosystems. We investigate the shared dynamics of spatially variable vegetation across three large watersheds in the southern Africa savanna. Dynamic Factor Analysis (DFA), a multivariate time-series dimension reduction technique, was used to identify the most important physical drivers of regional vegetation change. We first evaluated the Advanced Very High Resolution Radiometer (AVHRR)-vs. the Moderate Resolution Imaging Spectroradiometer (MODIS)-based Normalized Difference Vegetation Index (NDVI) datasets across their overlapping period (2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010). NDVI follows a general pattern of cyclic seasonal variation, with distinct spatio-temporal patterns across physio-geographic regions. Both NDVI products produced similar DFA models, although MODIS was simulated better. Soil moisture and precipitation controlled NDVI for mean annual precipitation (MAP) < 750 mm, and above this, evaporation and mean temperature dominated. A second DFA with the full AVHRR (1982-2010) data found that for OPEN ACCESSRemote Sens. 2013, 5 6514 MAP < 750 mm, soil moisture and actual evapotranspiration control NDVI dynamics, followed by mean and maximum temperatures. Above 950 mm, actual evapotranspiration and precipitation dominate. The quantification of the combined spatio-temporal environmental drivers of NDVI expands our ability to understand landscape level changes in vegetation evaluated through remote sensing and improves the basis for the management of vulnerable regions, like the southern Africa savannas.