Search

Addressing the global challenges of desertification, land degradation, and drought (DLDD), and their impacts on achieving sustainable development goals for coupled human-environmental systems is a key component of the 2030 Agenda for Sustainable Development. In particular, Sustainable Development Goal (SDG) 15.3 aims to, “by 2030, combat desertification, restore degraded land and soil, including land affected by desertification, drought and floods, and strive to achieve a land degradation-neutral world”. Addressing this challenge is essential for improving the livelihoods of those most affected by DLDD and for safeguarding against the most extreme effects of climate change. This paper introduces a conceptual framework for improved monitoring of DLDD in the context of United Nations Convention to Combat Desertification (UNCCD) Strategic Objective 2 (SO2) and its expected impacts: food security and adequate access to water for people in affected areas are improved; the livelihoods of people in affected areas are improved and diversified; local people, especially women and youth, are empowered and participate in decision-making processes in combating DLDD; and migration forced by desertification and land degradation is substantially reduced. While it is critical to develop methods and tools for assessing DLDD, work is needed first to provide a conceptual roadmap of the human dimensions of vulnerability in relation to DLDD, especially when attempting to create a globally standardized monitoring approach.

Despite sustained global efforts to avoid, reduce, and reverse land degradation, estimates of land degradation nationally and regionally vary considerably. Land degradation reduces agricultural productivity, impacts the provision of vital ecosystem services, and disproportionately affects vulnerable populations. The 2030 Agenda for Sustainable Development, through Sustainable Development Goal 15.3, sets out to achieve land degradation neutrality (LDN) by improving the livelihoods of those most affected and building resilience in areas affected by or at risk from degradation. The United Nations Convention to Combat Desertification (UNCCD) leads the charge in creating a spatially explicit framework for monitoring and reporting on LDN goals that countries can integrate into their land planning policies. However, it remains difficult to operationalize the integration of biophysical indicators of land degradation with climatic and socio-economic indicators to assess the impact of land degradation on vulnerable populations. We present an integrative framework that demonstrates how freely available global geospatial data sets can be leveraged through an open-source platform (Trends.Earth) to simplify and operationalize monitoring and reporting on progress towards achieving LDN. Then, we summarize a suite of data sets and approaches that can be used to understand and quantify the socio-ecological interactions between drought, land degradation and population exposed to desertification, land degradation and drought. We discuss how improvements in Earth observation data sets and algorithms will allow UNCCD land-based progress sub-indicators (changes in primary productivity, land cover, soil organic carbon, drought, and population exposure) to be computed at enhanced spatial resolutions.

Characterizing compositional and structural aspects of vegetation is critical to effectively assessing land function. When priorities are placed on ecological integrity, remotely sensed estimates of fractional vegetation components (FVCs) are useful for measuring landscape-level habitat structure and function. In this study, we address whether FVC estimates, stratified by dominant vegetation type, vary with different classification approaches applied to very-high-resolution small unoccupied aerial system (UAS)-derived imagery. Using Parrot Sequoia imagery, flown on a DJI Mavic Pro micro-quadcopter, we compare pixel- and segment-based random forest classifiers alongside a vegetation height-threshold model for characterizing the FVC in a southern African dryland savanna. Results show differences in agreement between each classification method, with the most disagreement in shrub-dominated sites. When compared to vegetation classes chosen by visual identification, the pixel-based random forest classifier had the highest overall agreement and was the only classifier not to differ significantly from the hand-delineated FVC estimation. However, when separating out woody biomass components of tree and shrub, the vegetation height-threshold performed better than both random-forest approaches. These findings underscore the utility and challenges represented by very-high-resolution multispectral UAS-derived data (~10 cm ground resolution) and their uses to estimate FVC. Semi-automated approaches statistically differ from by-hand estimation in most cases; however, we present insights for approaches that are applicable across varying vegetation types and structural conditions. Importantly, characterization of savanna land function cannot rely only on a “greenness” measure but also requires a structural vegetation component. Underscoring these insights is that the spatial heterogeneity of vegetation structure on the landscape broadly informs land management, from land allocation

Highland Andean ecosystems sustain high levels of floral and faunal biodiversity in areas with diverse topography and provide varied ecosystem services, including the supply of water to cities and downstream agricultural valleys. Google (™) has developed a product specifically designed for mapping purposes (Earth Engine), which enables users to harness the computing power of a cloud-based solution in near-real time for land cover change mapping and monitoring. We explore the feasibility of using this platform for mapping land cover types in topographically complex terrain with highly mixed vegetation types (Nor Yauyos Cochas Landscape Reserve located in the central Andes of Peru) using classification machine learning (ML) algorithms in combination with different sets of remote sensing data. The algorithms were trained using 3601 sampling pixels of (a) normalized spectral bands between the visible and near infrared spectrum of the Landsat 8 OLI sensor for the 2018 period, (b) spectral indices of vegetation, soil, water, snow, burned areas and bare ground and (c) topographic-derived indices (elevation, slope and aspect). Six ML algorithms were tested, including CART, random forest, gradient tree boosting, minimum distance, naïve Bayes and support vector machine. The results reveal that ML algorithms produce accurate classifications when spectral bands are used in conjunction with topographic indices, resulting in better discrimination among classes with similar spectral signatures such as pajonal (tussock grass-dominated cover) and short grasses or rocky groups, and moraines, agricultural and forested areas. The model with the highest explanatory power was obtained from the combination of spectral bands and topographic indices using the random forest algorithm (Kappa = 0.81). Our study presents a first approach of its kind in topographically complex Cordilleran terrain and we show that GEE is particularly useful in large-scale land cover mapping and monitoring in moun

During the 21st century, human-environment interactions will increasingly expose both systems to risks, but also yield opportunities for improvement as we gain insight into these complex, coupled systems. Human-environment interactions operate over multiple spatial and temporal scales, requiring large data volumes of multi-resolution information for analysis. Climate change, land-use change, urbanization, and wildfires, for example, can affect regions differently depending on ecological and socioeconomic structures. The relative scarcity of data on both humans and natural systems at the relevant extent can be prohibitive when pursuing inquiries into these complex relationships. We explore the value of multitemporal, high-density, and high-resolution LiDAR, imaging spectroscopy, and digital camera data from the National Ecological Observatory Network's Airborne Observation Platform (NEON AOP) for Socio-Environmental Systems (SES) research. In addition to providing an overview of NEON AOP datasets and outlining specific applications for addressing SES questions, we highlight current challenges and provide recommendations for the SES research community to improve and expand its use of this platform for SES research. The coordinated, nationwide AOP remote sensing data, collected annually over the next 30 yr, offer exciting opportunities for cross-site analyses and comparison, upscaling metrics derived from LiDAR and hyperspectral datasets across larger spatial extents, and addressing questions across diverse scales. Integrating AOP data with other SES datasets will allow researchers to investigate complex systems and provide urgently needed policy recommendations for socio-environmental challenges. We urge the SES research community to further explore questions and theories in social and economic disciplines that might leverage NEON AOP data.

During the 21st century, human–environment interactions will increasingly expose both systems to risks, but also yield opportunities for improvement as we gain insight into these complex, coupled systems. Human–environment interactions operate over multiple spatial and temporal scales, requiring large data volumes of multi-resolution information for analysis. Climate change, land-use change, urbanization, and wildfires, for example, can affect regions differently depending on ecological and socioeconomic structures. The relative scarcity of data on both humans and natural systems at the relevant extent can be prohibitive when pursuing inquiries into these complex relationships. We explore the value of multitemporal, high-density, and high-resolution LiDAR, imaging spectroscopy, and digital camera data from the National Ecological Observatory Network’s Airborne Observation Platform (NEON AOP) for Socio-Environmental Systems (SES) research. In addition to providing an overview of NEON AOP datasets and outlining specific applications for addressing SES questions, we highlight current challenges and provide recommendations for the SES research community to improve and expand its use of this platform for SES research. The coordinated, nationwide AOP remote sensing data, collected annually over the next 30 yr, offer exciting opportunities for cross-site analyses and comparison, upscaling metrics derived from LiDAR and hyperspectral datasets across larger spatial extents, and addressing questions across diverse scales. Integrating AOP data with other SES datasets will allow researchers to investigate complex systems and provide urgently needed policy recommendations for socio-environmental challenges. We urge the SES research community to further explore questions and theories in social and economic disciplines that might leverage NEON AOP data.

Remote sensing analyses focused on non-timber forest product (NTFP) collection and grazing are current research priorities of land systems science. However, mapping these particular land use patterns in rural heterogeneous landscapes is challenging because their potential signatures on the landscape cannot be positively identified without fine-scale land use data for validation. Using field-mapped resource areas and household survey data from participatory mapping research, we combined various Landsat-derived indices with ancillary data associated with human habitation to model the intensity of grazing and NTFP collection activities at 100-m spatial resolution. The study area is situated centrally within a transboundary southern African landscape that encompasses community-based organization (CBO) areas across three countries. We conducted four iterations of pixel-based random forest models, modifying the variable set to determine which of the covariates are most informative, using the best fit predictions to summarize and compare resource use intensity by resource type and across communities. Pixels within georeferenced, field-mapped resource areas were used as training data. All models had overall accuracies above 60% but those using proxies for human habitation were more robust, with overall accuracies above 90%. The contribution of Landsat data as utilized in our modeling framework was negligible, and further research must be conducted to extract greater value from Landsat or other optical remote sensing platforms to map these land use patterns at moderate resolution. We conclude that similar population proxy covariates should be included in future studies attempting to characterize communal resource use when traditional spectral signatures do not adequately capture resource use intensity alone. This study provides insights into modeling resource use activity when leveraging both remotely sensed data and proxies for human habitation in heterogeneous, spectrally