Search

Data Description These datasets were generated from the research article "The overlooked contribution of trees outside forests to tree cover and woody biomass across Europe". Here we publish aggregated version of canopy cover and height map, and originalbiomass map at 30m resolution. Thebiomass map are calculated from the canopy cover and height maps using allometric equations. The originalcanopy cover and height maps are derived from3m resolutionPlanetScope imagery (research license with Planet Labs Inc)in 2019, are available on request. Data Use Agreement The datasets are restricted to use only for research and scientific purposes, and shall not be used for any commercial purposes. Any usage beyond the permitted scope requires explicit written consent from the relevant parties. The restrictions using the datasets are coupled with the research license, which can be accessed here:https://assets.planet.com/docs/ToS_EducationAndResearch.pdf The user agrees to assume all risks and liabilities associated with the interpretation and use of any data or results obtained from the datasets. The University of Copenhagen, Planet Labs Inc, and any affiliated parties shall not be held liable for any consequences arising from the misuse of the data. &nbsp

Intelligent forest management is the key to mitigating climate warming, fostering a green economy, and protecting valuable habitats. Detailed knowledge about forests is a prerequisite for such management but is conventionally based on costly plot-scale data, rarely available at resolution of relevance for management strategies. Here, we present a deep learning-based framework that provides location, crown area and height for each individual tree from aerial images at country scale. We quantify and characterize all individual trees in Denmark and show that 26% of the trees grow outside forests, which is typically unrecognized in national inventories. Furthermore, we demonstrate that only marginal effort is needed to transfer our framework to Finland, despite markedly dissimilar landscapes and data sources. Our work lays the foundation for a global database, where every tree has its digital twin and is spatially traceable and manageable.

The consistent monitoring of trees both inside and outside of forests is key to mitigating climate change. However existing large-scale tree cover maps primarily quantify forest cover and do not include isolated trees, as these are not discernible in lower resolution satellite images. In many dryland countries these non-forest trees constitute the main form of tree cover, and play a vital role in ecological stability, local economies, livelihoods, and food security.Here we make use of the PlanetScope nanosatellite constellation, which delivers global very high-resolution daily imagery, to map both forest and non-forest tree cover for continental Africa using images from a single year. We composite 232,053 4-band scenes from 2019 into 1x1° mosaics and apply a Convolutional Neural Network to segment canopy cover of all trees and shrubs. To train the network we use a combination of manually annotated 1 m labels and source images upsampled from 3 m to 1 m, resulting in a final model that maps tree cover at 1 m across the continent, segmenting closed canopies in forest areas and individual scattered trees in savannah areas.Our prototype map demonstrates that a precise assessment of all tree-based ecosystems is possible at continental scale, and reveals that 29% of tree cover is found outside areas previously classified as tree cover in state-of-the-art maps, such as in croplands and grassland. This analysis lays the groundwork towards global scale studies of tree cover at individual tree level and annual temporal scale, which is crucial for improved managing of woody resources, monitoring of TOF in relation to agroforestry, tree planting and restoration efforts, and assessing land use impacts in non-forest landscapes.

Distinguishing trees on agricultural land from forests is essential for a better understanding of the relationship between forests and human farming activities. However, it is difficult to separate them with remote sensing imagery since they share similar canopy cover, especially on the edge of the amazon rain forest, which has a much-complicated agriculture pattern. Except for annual crops and pasture, there are also lots of agroforestry applications and shifting cultivation, which integrates many tree systems. And those tree systems are not well separated from the forest in the existing land cover map. Recent techniques allow for the mapping of single trees outside of forests, now we take the next step by identifying those diverse tree-involved systems in agricultural land. Here we aim to generate a robust, cost-efficient method to distinguish trees within agricultural land from the forest. We started our exploration from Peruvian Amazon, where the competition for land has increased in the last decades, causing possible adverse effects on livelihoods and ecosystem services. Deep learning models, data sampling, and fine-tuning strategies are tested and optimized with PlanetScope satellite imagery. Our research target is to provide a tool for separating tree systems in farmland from the forest. It can also be used as a base map to explore the dynamic of agriculture transition and its impact on livelihoods and ecosystem services.

Trees sustain livelihoods and mitigate climate change but a predominance of trees outside forests and limited resources make it difficult for many tropical countries to conduct automated nation-wide inventories. Here, we propose an approach to map the carbon stock of each individual overstory tree at the national scale of Rwanda using aerial imagery from 2008 and deep learning. We show that 72% of the mapped trees are located in farmlands and savannas and 17% in plantations, accounting for 48.6% of the national aboveground carbon stocks. Natural forests cover 11% of the total tree count and 51.4% of the national carbon stocks, with an overall carbon stock uncertainty of 16.9%. The mapping of all trees allows partitioning to any landscapes classification and is urgently needed for effective planning and monitoring of restoration activities as well as for optimization of carbon sequestration, biodiversity and economic benefits of trees. ispartof: NATURE CLIMATE CHANGE vol:13 issue:1 pages:91-+ ispartof: location:England status: published

Sustainable tree resource management is the key to mitigating climate warming, fostering a green economy, and protecting valuable habitats. Detailed knowledge about tree resources is a prerequisite for such management but is conventionally based on plot-scale data, which often neglects trees outside forests. Here, we present a deep learning-based framework that provides location, crown area, and height for individual overstory trees from aerial images at country scale. We apply the framework on data covering Denmark and show that large trees (stem diameter >10 cm) can be identified with a low bias (12.5%) and that trees outside forests contribute to 30% of the total tree cover, which is typically unrecognized in national inventories. The bias is high (46.6%) when our results are evaluated against all trees taller than 1.3 m, which involve undetectable small or understory trees. Furthermore, we demonstrate that only marginal effort is needed to transfer our framework to data from Finland, despite markedly dissimilar data sources. Our work lays the foundation for digitalized national databases, where large trees are spatially traceable and manageable.

Wetlands provide invaluable services for ecosystems and society and are a crucial instrument in our fight against climate change. Although Earth Observation satellites offer cost-effective and accurate information about wetland status at the continental scale; to date, there is no universally accepted, standardized, and regularly updated inventory of European wetlands 85%) that a substantial extent of European peatlands was previously classified as grassland and other land cover types. In addition, our map highlights cultivated areas (e.g., river floodplains) that can be potentially rewetted. Such accurate and consistent mapping of different wetland types at a continental scale offers a baseline for future wetland monitoring and trend assessment, supports the detailed reporting of European carbon budgets, and lays down the foundation towards a global wetland inventory.

Transitioning from the expansion of croplands to sustainable agricultural intensification is key to ensuring food and nutrition security. This is particularly the case for small-scale subsistence farmers in drylands, where our current understanding on the sustainability of this transition is however mostly derived from theories based on local observations. Here we demonstrate that a new generation of satellite data and computer science technology can be used to set a new baseline to understand the interplay between population pressure and agricultural intensification. By characterizing cropping systems across the African Sahel, we evidence that agroforestry areas of high land use intensity are associated with both high rural population density and crop growth rate, upholding Boserup’s theory of agricultural intensification. Our approach provides the foundation for sustainable agricultural intensification in drylands and represents a pathway towards the integration of multidimensional cropland maps into context-specific policymaking for low- and middle- income countries.

The consistent monitoring of trees both inside and outside of forests is key to mitigating climate change. Current monitoring systems either ignore trees outside forests or are too expensive to be applied consistently across countries on a repeated basis. Here we make use of the PlanetScope nanosatellite constellation, which delivers global very high-resolution daily imagery, to map both forest and non-forest tree cover for continental Africa using images from a single year. Our prototype map of 2019 demonstrates that a precise assessment of all tree-based systems is possible at continental scale, and reveals that 29% of tree cover is found outside areas previously classified as tree cover, such as in croplands and grassland. Such accurate mapping of tree cover at metric resolution down to the level of individual trees and consistent among countries has the potential to redefine land use impacts, move beyond the need for forest definitions, build the basis for natural climate solutions, and provide a new scientific basis for tree related studies.

The consistent monitoring of trees both inside and outside of forests is key to mitigating climate change. Current monitoring systems either ignore trees outside forests or are too expensive to be applied consistently across countries on a repeated basis. Here we make use of the PlanetScope nanosatellite constellation, which delivers global very high-resolution daily imagery, to map both forest and non-forest tree cover for continental Africa using images from a single year. Our prototype map of 2019 demonstrates that a precise assessment of all tree-based systems is possible at continental scale, and reveals that 29% of tree cover is found outside areas previously classified as tree cover, such as in croplands and grassland. Such accurate mapping of tree cover at metric resolution down to the level of individual trees and consistent among countries has the potential to redefine land use impacts, move beyond the need for forest definitions, build the basis for natural climate solutions, and provide a new scientific basis for tree related studies.

Earth Observation (EO) data is a critical information source for mapping and monitoring water resources over large inaccessible regions where hydrological in-situ networks are sparse. In this paper, we present a simple yet robust method for fusing optical and Synthetic Aperture Radar (SAR) data for mapping surface water dynamics over mainland China. This method uses a multivariate logistic regression model to estimate monthly surface water extent over a four-year period (2017 to 2020) from the combined usages of Sentinel-1, Sentinel-2 and Landsat-8 imagery. Multi-seasonal high-resolution images from the Chinese Gaofen satellites are used as a reference for an independent validation showing a high degree of agreement (overall accuracy 94%) across a diversity of climatic and physiographic regions demonstrating potential scalability beyond China. Through inter-comparison with similar global scale products, this paper further shows how this new mapping technique provides improved spatio-temporal characterization of inland water bodies, and for better capturing smaller water bodies (< 0.81 ha in size). The relevance of the results is discussed, and we find this new enhanced monitoring approach has the potential to advance the use of Earth observation for water resource management, planning and reporting.

International audience; Vegetation phenology obtained from time series of remote sensing data is relevant for a range of ecological applications. The freely available Sentinel-2 imagery at a 10 m spatial resolution with a similar to 5-day repeat cycle provides an opportunity to map vegetation phenology at an unprecedented fine spatial scale. To facilitate the production of a Europe-wide Copernicus Land Monitoring Sentinel-2 based phenology dataset, we design and evaluate a framework based on a comprehensive set of ground observations, including eddy covariance gross primary production (GPP), PhenoCam green chromatic coordinate (GCC), and phenology phases from the Pan-European Phenological database (PEP725). We test three vegetation indices (VI) - the normalized difference vegetation index (NDVI), the two-band enhanced vegetation index (EVI2), and the plant phenology index (PPI) - regarding their capability to track the seasonal trajectories of GPP and GCC and their performance in reflecting spatial variabilities of the corresponding GPP and GCC phenometrics, i.e., start of season (SOS) and end of season (EOS). We find that for GPP phenology, PPI performs the best, in particular for evergreen coniferous forest areas where the seasonal variations in leaf area are small and snow is prevalent during wintertime. Results are inconclusive for GCC phenology, for which no index is consistently better than the others. When comparing to PEP725 phenology phases, PPI and EVI2 perform better than NDVI regarding the spatial correlation and consistency (i.e., lower standard deviation). We also link VI phenometrics at various amplitude thresholds to the PEP725 phenophases and find that PPI SOS at 25% and PPI EOS at 15% provide the best matches with the ground-observed phenological stages. Finally, we demonstrate that applying bidirectional reflectance distribution function correction to Sentinel-2 reflectance is a step that can be excluded for phenology mapping in Europe.

Remote sensing-derived cropland products have depicted the location and extent of agricultural lands with an ever increasing accuracy. However, limited attention has been devoted to distinguishing between actively cropped fields and fallowed fields within agricultural lands, and in particular so in grass fallow systems of semi-arid areas. In the Sahel, one of the largest dryland regions worldwide, crop-fallow rotation practices are widely used for soil fertility regeneration. Yet, little is known about the extent of fallow fields since fallow is not explicitly differentiated within the cropland class in any existing remote sensing-based land use/cover maps, regardless of the spatial scale. With a 10 m spatial resolution and a 5-day revisit frequency, Sentinel-2 satellite imagery made it possible to disentangle agricultural land into cropped and fallow fields, facilitated by Google Earth Engine (GEE) for big data handling. Here we produce the first Sahelian fallow field map at a 10 m resolution for the baseline year 2017, accomplished by designing a remote sensing driven protocol for generating reference data for mapping over large areas. Based on the 2015 Copernicus Dynamic Land Cover map at 100 m resolution, the extent of fallow fields in the cropland class is estimated to be 63% (403,617 km2) for the Sahel in 2017. Similar results are obtained for five contemporary cropland products, with fallow fields occupying 57–62% of the cropland area. Yet, it is noted that the total estimated area coverage depends on the quality of the different cropland products. The share of cropped fields within the Copernicus cropland area is found to be higher in the arid regions (200–300 mm rainfall) as compared to the semi-arid regions (300–600 mm rainfall). The woody cover fraction within cropped and fallow fields is found to have a reversed pattern between arid (higher woody cover in cropped fields) and semi-arid (higher woody cover in fallow fields) regions. The method developed, using cloud-based Earth Observation (EO) data and computation on the GEE platform, is expected to be reproducible for mapping the extent of fallow fields across global croplands. Future applications based on multi-year time series is expected to improve our understanding of crop-fallow rotation dynamics in grass fallow systems being key in teasing apart how cropland intensification and expansion affect environmental variables, such as soil fertility, crop yields and local livelihoods in low-income regions such as the Sahel. The mapping result can be visualized via a web viewer (https://buwuyou.users.earthengine.app/view/fallowinsahel).

It has been widely asserted that a high grazing pressure has led to a reduction in vegetation production at decadal time scales, implying land degradation, in African drylands, and in the Sahel in particular. We test this hypothesis by analyzing spatiotemporal patterns of vegetation production in the northwestern Ferlo in Senegal. Normalized difference vegetation index (NDVI) patterns, here used to represent vegetation production, were determined on the basis of Earth observation MODIS and PROBA‐V data. Furthermore, woody cover was assessed by very high spatial resolution (VHR) imagery. Because livestock is concentrated around deep wells in the dry season for watering, while for management they concentrate daily around pastoral family settlements or camps all year round, we studied the dependence of NDVI on distance from deep wells and camps. Locations of wells and camps were determined by visual inspection of VHR images. We found that 'inverse grazing gradients,' defined as decreasing NDVI/production with increasing distance, dominated both around wells and camps. Further, we found neither decrease nor increase in woody cover with distance to wells. Both positive and negative trends in wet season NDVI over the period 2000–2016 were identified within the study area, yet temporal trends were predominantly positive in proximity to deep wells. Around pastoral camps, positive NDVI trends were generally elevated relative to the regional average. The results question the validity of claims that high grazing pressure causes land degradation at the landscape scale, yet they show that grazing does lead to substantial spatial redistribution of vegetation production.

Tropical savannas comprise mixed woodland grassland ecosystems in which trees and grasses compete for water resources thereby maintaining the spatial structuring of this ecosystem. A global change in rainfall climatology may impact the structure of tropical savanna ecosystems by favouring woody plants, relative to herbaceous vegetation. Here we analysed satellite data and observed a relatively higher increase in woody vegetation (5%) as compared to the increase in annual maximum leaf area index (LAImax, an indicator of the total green vegetation production) (3%) in arid and semi-arid savannas over recent decades. We further observed a declining sensitivity of LAImax to annual rainfall over 56% of the tropical savannas, spatially overlapping with areas of increased woody cover and altered rainfall climatology. This suggests a climate-induced shift in the coexistence of woody and herbaceous vegetation in savanna ecosystems, possibly caused by altered hydrological conditions with significance for land cover and associated biophysical effects such as surface albedo and evapotranspiration., Changing rainfall patterns may drive changes in the structure of tropical savanna. Here Zhang et al. use satellite data from global tropical savannas, and find evidence to suggest that altered rainfall may be favouring woody plants over herbaceous plants in these ecosystems.

Woody vegetation in farmland acts as a carbon sink and provides ecosystem services for local people, but no macroscale assessments of the impact of management and climate on woody cover exist for drylands. Here we make use of very high spatial resolution satellite imagery to derive wall-to-wall woody cover patterns in tropical West African drylands. Our study reveals that mean woody cover in farmlands along all semi-arid and sub-humid rainfall zones is 16%, on average only 6% lower than in savannahs. In semi-arid Sahel, farmland management promotes woody cover around villages (11%), while neighbouring savannahs had on average less woody cover. However, farmlands in sub-humid zones have a greatly reduced woody cover (21%) as compared with savannahs (33%). In the region as a whole, rainfall, terrain and soil are the most important (80%) determinants of woody cover, while management factors play a smaller (20%) role. We conclude that agricultural expansion causes a considerable reduction of trees in woodlands, but observations in Sahel indicate that villagers safeguard trees on nearby farmlands which contradicts simplistic ideas of a high negative correlation between population density and woody cover.

Climate change in drylands has caused alterations in the seasonal distribution of rainfall including increased heavy-rainfall events, longer dry spells, and a shifted timing of the wet season. Yet the aboveground net primary productivity (ANPP) in drylands is usually explained by annual-rainfall sums, disregarding the influence of the seasonal distribution of rainfall. This study tested the importance of rainfall metrics in the wet season (onset and cessation of the wet season, number of rainy days, rainfall intensity, number of consecutive dry days, and heavy-rainfall events) for growing season ANPP. We focused on the Sahel and northern Sudanian region (100–800 mm yr−1) and applied daily satellite-based rainfall estimates (CHIRPS v2.0) and growing-season-integrated normalized difference vegetation index (NDVI; MODIS) as a proxy for ANPP over the study period: 2001–2015. Growing season ANPP in the arid zone (100–300 mm yr−1) was found to be rather insensitive to variations in the seasonal-rainfall metrics, whereas vegetation in the semi-arid zone (300–700 mm yr−1) was significantly impacted by most metrics, especially by the number of rainy days and timing (onset and cessation) of the wet season. We analysed critical breakpoints for all metrics to test if vegetation response to changes in a given rainfall metric surpasses a threshold beyond which vegetation functioning is significantly altered. It was shown that growing season ANPP was particularly negatively impacted after >14 consecutive dry days and that a rainfall intensity of ∼13 mm day−1 was detected for optimum growing season ANPP. We conclude that the number of rainy days and the timing of the wet season are seasonal-rainfall metrics that are decisive for favourable vegetation growth in the semi-arid Sahel and need to be considered when modelling primary productivity from rainfall in the drylands of the Sahel and elsewhere.

International audience; Changing woody plant density across agricultural landscapes of the Sudano-Sahelian region is a debated issue. This paper reports the results of an investigation on the contribution of field hedges to overall woody plant density. Hedges separating individual cropped parcels were studied within village agropastoral territories in the Dantiandou district in western Niger. In 1992, field hedges were mapped over the study area using aerial photography and in 2016, using high resolution Google Earth imagery. In 1992, field hedge length was equal to 1006 km within 435 km 2 of croplands, equivalent to 2.3 km km −2. In 2016, 17.5% of these hedges had disappeared , but 1591 km of new hedges were observed resulting in an increased density of hedges to 5.6 km km −2. In 24 years, hedges had increased at a mean annual rate of 3.7% likely associated with the splitting of crop fields by inheritance. The composition and productivity of hedges were also monitored in 1996, 2010 and 2015. All trees (max-imum height ≥ 4 m), shrubs (< 4 m) and tussock perennial grasses were recorded within twelve field hedge samples of 200 m each. Field measurements were used to estimate basal area, crown area, foliage mass and wood mass of each woody plant within the sampled hedges. No significant trends were found between 1996 and 2015 in woody plant density, basal area, crown cover, wood and foliage masses across all monitored hedges. However, overall means hide a slight decrease in tree contributions, while shrub contributions first increased and then decreased. They also mask contradictory trends among sites most likely related to different rates of shrub coppicing and tree cutting. The woody species composition of the hedges is poor with an increase of Combretum glutinosum to the detriment of Guiera senegalensis over the study period. Multiplying the hedge sample statistics by the changing lengths of field hedges in the study area provides estimates of the contribution of the hedge woody plants to the woody plant population at a landscape scale. Between 1992 and 2016, field hedges contributed to increases of woody plant density by 3.9%, basal areas by 5.4%, crown cover by 2.7%, leading to 6.1% annual increase in foliage mass and 8.8% increase in wood mass.

Changes in vegetation phenology are among the most sensitive biological responses to global change. While land surface phenological changes in the Northern Hemisphere have been extensively studied from the widely used long-term AVHRR (Advanced Very High Resolution Radiometer) data, current knowledge on land surface phenological trends and the associated drivers remains uncertain for the tropics. This uncertainty is partly due to the well-known challenges of applying satellite-derived vegetation indices from the optical domain in areas prone to frequent cloud cover. The long-term vegetation optical depth (VOD) product from satellite passive microwaves features less sensitivity to atmospheric perturbations and measures different vegetation traits and functioning as compared to optical sensors. VOD thereby provides an independent and complementary data source for studying land surface phenology and here we performed a combined analysis of the VOD and AVHRR NDVI (Normalized Difference Vegetation Index) datasets for the dry tropics (25°N to 25°S) during 1992–2012. We find a general delay in the VOD derived start of season (SOS) and end of season (EOS) as compared to NDVI derived metrics, however with clear differences among land cover and continents. Pixels characterized by significant phenological trends (P < 0.05) account for up to 20% of the study area for each phenological metric of NDVI and VOD, with large spatial difference between the two sensor systems. About 50% of the pixels studied show significant phenological changes in either VOD or NDVI metrics. Drivers of phenological changes were assessed for pixels of high agreement between VOD and NDVI phenological metrics (serving as a means of reducing noise-related uncertainty). We find rainfall variability and woody vegetation change to be the main forcing variables of phenological trends for most of the dry tropical biomes, while fire events and land cover change are recognized as second-order drivers. Taken together, our study provides new insights on land surface phenological changes and the associated drivers in the dry tropics, as based on the complementary long-term data sources of VOD and NDVI, sensitive to changes in vegetation water content and greenness, respectively.

International audience; Plant water storage is fundamental to the functioning of terrestrial ecosystems by participating in plant metabolism, nutrient and sugar transport, and maintenance of the integrity of the hydraulic system of the plant. However, a global view of the size and dynamics of the water pools stored in plant tissues is still lacking. Here, we report global patterns of seasonal variations in ecosystem-scale plant water storage and their relationship with leaf phenology, based on space-borne measurements of L-band vegetation optical depth. We find that seasonal variations in plant water storage are highly synchronous with leaf phenology for the boreal and temperate forests, but asynchronous for the tropical woodlands, where the seasonal development of plant water storage lags behind leaf area by up to 180 days. Contrasting patterns of the time lag between plant water storage and terrestrial groundwater storage are also evident in these ecosystems. A comparison of the water cycle components in seasonally dry tropical woodlands highlights the buffering effect of plant water storage on the seasonal dynamics of water supply and demand. Our results offer insights into ecosystem-scale plant water relations globally and provide a basis for an improved parameterization of eco-hydrological and Earth system models.

The impact of human activities via land use/cover changes on NDVI trends is critical for an improved understanding of satellite-observed changes in vegetation productivity in drylands. The dominance of positive NDVI trends in the Sahel, the so-called re-greening, is sometimes interpreted as a combined effect of an increase in rainfall and cropland expansion or agricultural intensification. Yet, the impact of changes in land use has yet to be thoroughly tested and supported by empirical evidence. At present, no studies have considered the importance of the different seasonal NDVI signals of cropped and fallowed fields when interpreting NDVI trends, as both field types are commonly merged into a single ‘cropland’ class. We make use of the distinctly different phenology of cropped and fallowed fields and use seasonal NDVI curves to separate these two field types. A fuzzy classifier is applied to quantify cropped and fallowed areas in a case study region in the southern Sahel (Fakara, Niger) on a yearly basis between 2000 and 2014. We find that fallowed fields have a consistently higher NDVI than unmanured cropped fields and by using two seasonal NDVI metrics (the amplitude and the decreasing rate) derived from the MODIS time series, a clear separation between classes of fields is achieved (r = 0.77). The fuzzy classifier can compute the percentage of a pixel (250 m) under active cultivation, thereby alleviating the problem of small field sizes in the region. We find a predominant decrease in NDVI over the period of analysis associated with an increased area of cropped fields at the expense of fallowed fields. Our findings couple cropping abandonment (more frequent fallow years) with positive NDVI trends and an increase in the percentage of the cropped area (fallow period shortening) with negative trends. These findings profoundly impact our understanding of greening and browning trends in agrarian Sahelian drylands and in other drylands of developing countries characterized by limited use of fertilizers.

Woody vegetation in farmland acts as a carbon sink and provides ecosystem services for local people, but no macro-scale assessments of the impact of management and climate on woody cover exists for drylands. Here we make use of very high spatial resolution satellite imagery to derive wall-to-wall woody cover patterns in tropical West African drylands. In arid and semi-arid Sahel, areas of more people are associated with more trees: mean woody cover is greater in farmlands (12%) than in savannas (6%), and likewise it is higher close to villages than further away. In sub-humid savannas of West Africa, woody cover is generally above 20% and decreases with increasing population density, but remains around 15% in farmlands, independent of rainfall. In the region as a whole, rainfall, terrain and soil are the most important (80%) determinants of woody cover, while management factors play a smaller (20%) role. We conclude that agricultural expansion cannot generally be claimed to cause woody cover losses, and that observations in Sahel contradict simplistic ideas of a high negative correlation between population density and woody cover.

Spatial distribution information on population density is essential for understanding urban dynamics. In recent decades, remote sensing techniques have often been applied to assess population density, particularly night-time light data (NTL). However, such attempts have resulted in mapped population density at coarse/medium resolution, which often limits the applicability of such data for fine-scale territorial planning. The improved quality and availability of multi-source remote sensing imagery and location-based service data (LBS) (from mobile networks or social media) offers new potential for providing more accurate population information at the micro-scale level. In this paper, we developed a fine-scale population distribution mapping approach by combining the functional zones (FZ) mapped with high-resolution satellite images, NTL data, and LBS data. Considering the possible variations in the relationship between population distribution and nightlight brightness in functional zones, we tested and found spatial heterogeneity of the relationship between NTL and the population density of LBS samples. Geographically weighted regression (GWR) was thus implemented to test potential improvements to the mapping accuracy. The performance of the following four models was evaluated: only ordinary least squares regression (OLS), only GWR, OLS with functional zones (OLS&FZ) and GWR with functional zones (GWR&FZ). The results showed that NTL-based GWR&FZ was the most accurate and robust approach, with an accuracy of 0.71, while the mapped population density was at a unit of 30 m spatial resolution. The detailed population density maps developed in our approach can contribute to fine-scale urban planning, healthcare and emergency responses in many parts of the world.