The occurrence of natural vegetation at a given time is determined by interplay of multiple drivers. The effects of several drivers, e.g., geomorphology, topography, climate variability, accessibility, demographic indicators, and changes in human activities on the occurrence of natural vegetation in the severe drought periods and, prior to the year 2000, have been analyzed in West Africa. A binary logistic regression (BLR) model was developed to better understand whether the variability in these drivers over the past years was statistically significant in explaining the occurrence of natural vegetation in the year 2000. Our results showed that multiple drivers explained the occurrence of natural vegetation in West Africa at p < 0.05. The dominant drivers, however, were site-specific. Overall, human influence indicators were the dominant drivers in explaining the occurrence of natural vegetation in the selected hotspots. Human appropriation of net primary productivity (HANPP), which is an indicator of human socio-economic activities, explained the decreased likelihood of natural vegetation occurrence at all the study sites. However, the impacts of the remaining significant drivers on natural vegetation were either positive (increased the probability of occurrence) or negative (decreased the probability of occurrence), depending on the unique environmental and socio-economic conditions of the areas under consideration. The study highlights the significant role human activities play in altering the normal functioning of the ecosystem by means of a statistical model. The research contributes to a better understanding of the relationships and the interactions between multiple drivers and the response of natural vegetation in West Africa. The results are likely to be useful for planning climate change adaptation and sustainable development programs in West Africa.@en

Post‐classification change detection was applied to examine the nature of Land Use Land Cover (LULC) transitions in West Africa in three time intervals (1975–2000, 2000–2013, and 1975– 2013). Detailed analyses at hotspots coupled with comparison of LULC transitions in the humid and arid regions were undertaken. Climate and anthropic drivers of environmental change were disentangled by the LULC transitions analyses. The results indicated that human‐managed LULC types have replaced the natural LULC types. The total vegetation cover declined by −1.6%. Massive net gains in croplands (107.8%) and settlements (140%) at the expense of natural vegetation were detected in the entire period (1975–2013). Settlements expanded in parallel with cropland, which suggests the effort to increase food production to support the increasing population. Expansion of artificial water bodies were detected in the humid regions during the period of 1975–2000. Nonetheless, shrinking of water bodies due to encroachment by wetlands and other vegetation was observed in the arid regions, coupled with net loss in the whole of West Africa. The results indicate deforestation and degradation of natural vegetation and water resources in West Africa. Underlying anthropic drivers and a combination of anthropic and climate drivers were detected. LULC transitions in West Africa are location specific and have both positive and negative implications on the environment. The transitions indicate how processes at the local level, driven by human activities, lead to changes at the continental level and may contribute to global environmental change.@en

We present in this paper a polynomial fitting method applicable to segments of footprints measured by the Geoscience Laser Altimeter System (GLAS) to estimate glacier thickness change. Our modification makes the method applicable to complex topography, such as a large mountain glacier. After a full analysis of the planar fitting method to characterize errors of estimates due to complex topography, we developed an improved fitting method by adjusting a binary polynomial surface to local topography. The improved method and the planar fitting method were tested on the accumulation areas of the Naimona’nyi glacier and Yanong glacier on along-track facets with lengths of 1000 m, 1500 m, 2000 m, and 2500 m, respectively. The results show that the improved method gives more reliable estimates of changes in elevation than planar fitting. The improved method was also tested on Guliya glacier with a large and relatively flat area and the Chasku Muba glacier with very complex topography. The results in these test sites demonstrate that the improved method can give estimates of glacier thickness change on glaciers with a large area and a complex topography. Additionally, the improved method based on GLAS Data and Shuttle Radar Topography Mission-Digital Elevation Model (SRTM-DEM) can give estimates of glacier thickness change from 2000 to 2008/2009, since it takes the 2000 SRTM-DEM as a reference, which is a longer period than 2004 to 2008/2009, when using the GLAS data only and the planar fitting method.@en

In this paper, we propose a novel approach to video captioning based on adversarial learning and long short-term memory (LSTM). With this solution concept, we aim at compensating for the deficiencies of LSTM-based video captioning methods that generally show potential to effectively handle temporal nature of video data when generating captions but also typically suffer from exponential error accumulation. Specifically, we adopt a standard generative adversarial network (GAN) architecture, characterized by an interplay of two competing processes: a 'generator' that generates textual sentences given the visual content of a video and a 'discriminator' that controls the accuracy of the generated sentences. The discriminator acts as an 'adversary' toward the generator, and with its controlling mechanism, it helps the generator to become more accurate. For the generator module, we take an existing video captioning concept using LSTM network. For the discriminator, we propose a novel realization specifically tuned for the video captioning problem and taking both the sentences and video features as input. This leads to our proposed LSTM-GAN system architecture, for which we show experimentally to significantly outperform the existing methods on standard public datasets.@en