Inventorying tree intensification to quantify shade/canopy cover variability in multi-strata tropical agroforestry land use(s) - Case of Cocoa Agroforestry in Cameroon
In the forested south of Cameroon, Cocoa Agroforestry is an important land use and livelihood strategy. A growing demand for Cocoa and by-products has sparked a recent boost in governmental strategies to increase production and national export earnings, through intensifying management and extensive land use. This land use is practised under a multi-strata shade/canopy system; with a contending shade tree management (retention and integration) varying with site characteristics, farm management objectives and farmers’ experience. In such system, shade provision is an important determinant of production capacity and environmental resilience – already evidence as reforestation option in savannah-forest transition of central Cameroon. There are ample reports in literature revealing the economic importance and need of tree integration within the system, while a few studies have been pivotal in ascertaining their ecosystem services (ES) potential as such as biodiversity and carbon storage. Yet lacking is research-based information on shade cover variation with the tree management strategy employed at temporal scale. Such empirical evidence is much needed to support sustainable management and livelihood options.
Measurements from spaceborne Earth Observation satellites offer efficient, less laborious and large-scale assessment of agricultural vegetation. Such measurements have proved successful in assessing tree attributes within savannah or Sahel agroforests, using OBIA approaches. In tropical agroforestry conditions, issues of mixed features/trees and variable clusters of vegetation structures present challenges in multi-strata canopy assessment, and thus require customized image analysis protocols.
Sustainable multiple-use management of Colophospermum mopane woodlands in northern Namibia
In the northern part of Namibia, mopane (a xeric, multi-purpose tree species) woodland covers 10 million ha and provides many goods and services to rural communities who live near populations of this species. Recently however there is a declining trend in the mopane woodland due to demographic, economic, environmental and cultural changes. New policies and approaches are being adopted and management processes are being adjusted. There have been a number of valuable studies about decision making and management planning related to multiple use and sustainability of mopane woodlands in particular. However, following research gaps were identified:
Lack of integration of research findings and local knowledge in woodland management
Lack of quantitative information about NTFPs and their integration in woodland management
Lack of integration of socio-economic data
The aim of the research presented here is to contribute to the development of the necessary scientific basis for sustainable multiple-use woodland management in the semi-arid woodland region in general and in the mopane woodlands in northern Namibia particularly.
Impact of commercialization on the population parameters of Irvingia, a Non Timber Forest Products (NTFP) genus in Cameroon
Non-timber forest products (NTFPs) have the potential to alleviate poverty in combination with nature conservation. Hence, the commercialization of NTFPs has been widely promoted. Nevertheless, the actual impact of promoting and increasing NTFP commercialization on the environment and livelihoods of farmers has barely been assessed. Against this background, this study assessed the promotion of Ricinodendron heudelotii (Baill.) Pierre ex Pax. kernel (njansang) commercialization by the World Agroforestry Centre (ICRAF) in project villages in Cameroon with the aim to alleviate poverty for small-scale farmers. The 5-year impact of development interventions on farmers livelihoods were investigated, comparing project and control households.. A five pillar framework was applied, tackling the financial, social, human, physical and natural assets of farmers.
ENDELEO: Development of a remote sensing derived tool to assess the impact of conservation policy measures and drought on East African ecosystems
The ENDELEO project aims at the promotion of good environmental governance in East Africa of fragile ecosystems through remote sensing. It focuses on two inter-related areas (1) Enhancing the understanding of management issues in drought vulnerable areas, in particular arid and semi-arid areas, and (2) providing concerned stakeholders with remote sensing based information tools to foster better environmental governance.
The sustainability of the monitoring system is ensured through complementing on past and on-going efforts in understanding the state of, and the trends in conservation of major ecosystems in East-Africa, and the involvement of stakeholders at all stages of the project.
ENDELEO will fully support UNEP planned 10-year action plan for environmental recovery and measures for addressing vulnerability to drought, including policies, legislation and strategies in the Horn of Africa.
Monitoring and modeling of dynamical wetland processes in the Sahel region using time series of optical satellite imagery and climate data
The Sahel region in West-Africa contains a large number of wetlands, mainly floodplains. In an average year the total inundated area of the major wetland systems in the Sahel is about 67,000 km. These ecosystems are of high ecological value. Several wetlands, such as the Inner Niger Delta, are included in the Ramsar list. The Sahelian wetlands also have a high economical value. Local communities make use of the floodplains for agriculture, dry season grazing and fishing.
Sustainable management of the Sahelian wetlands requires a balance between ecological and economical functions. This can be facilitated by mapping and monitoring of these natural resources . Due to the extent and inaccessibility of the floodplains, the only possible way to do this on a large scale is by means of remote sensing. Monitoring can be performed using a time series of satellite data to observe changes in size and composition.
Reusing Artificial Neural Networks to enhance satellite image land cover classification
The use of artificial neural networks (ANNs) within the context of remote sensing has largely increased. Mostly, ANNs are used as a classifier, classifying every unknown pixel in a possibly multi- band image in land cover / land use classes. Compared to more conventional classifiers like the maximum likelihood classifier, ANNs feature several advantages: they are considered as efficient, are non-parametric, and are relatively resistant to noise. Still, ANNs are often considered black-box models with poor semantic qualities. Furthermore, training (calibration) requires a certain amount of training data and the tuning of the network architecture and learning parameters is often a time consuming trial and error process.
Literature often suggests rules of thumb to address these problems. These heuristics are often contradictionary, and they never guarantee a successful training process. This project therefor aims at developing and applying a method that overcomes the aforementioned problems.
Imaging radar for mapping and monitoring of wetland ecosystems in the Lake Chad Basin
This research focuses on Sahelian wetlands, which are amongst the world's largest swamps. These have been put under increasing pressure during the last 40 years due to the growth of the human population and the associated increased demand for irrigation water and arable land. In an effort to improve the economic situation of the local populations, large-scale hydro-agricultural projects are being planned, comprising large dams and extensive irrigation schemes. These dams have a negative impact on the wetlands endangering the local economies and regional biodiversity. To assess the impact of these measurements remote sensing data is used to supply the information needed to produce land cover maps, which are often not available in developing countries. Satellite imagery from different sources has been acquired, which are combined using digital image fusion protocols. The aim is to produce land cover maps covering large areas retaining a relatively high spatial resolution.
Mapping and modelling of temporal and bio-physical phenomena of the Sahelian wetlands environment using VEGETATION data
The Sahelian zone supports some of the world's largest swamps. During the last 40 years the growth of the human population and the associated increased demand for irrigation water and arable land, has put increasing pressure on these wetland ecosystems. In an effort to improve the economic situation of the local populations, large-scale hydro-agricultural projects are being build, comprising large dams and extensive irrigation schemes. The vegetation of the wetlands is influenced by two phenomena:
the alternation between dry and wet season
the filling and emptying of the floodplains
These characteristics make the wetlands suitable objects for study using satellite images with coarse spatial but high temporal resolution. As the wetlands of the Lake Chad basin have been the focuses of several conservation efforts during the last decade, the current project focuses on mapping and modelling of temporal and bio-physical phenomena in this region.
Development of Sustainability Assessment in Rural Multiple Use Forest in the Upper Mekong Watershed, Northwest Yunnan, China
In most developing countries, the livelihoods of rural people highly depend upon natural resources. They typically use the forest based on traditional forest knowledge (TFK). Meanwhile there is as yet few research-based evidence that enables to determine the constraints to forest sustainability induced by multiple uses by indigenous people at local level. On the other hand, it is of paramount importance to integrate TFK and modern scientific forest management methods into feasible sustainable forest management (SFM).
Northwest Yunnan is well known in China for its biological and cultural diversity. However, the mountain societies feature a significant degree of poverty. In this research project, a new paradigm of SFM for poor and natural resource dependent communities in the mountain region is explored. The aim is to chart the constraints to local use and management and to develop methods oriented towards science based eco-management of local forest resources.
Tourism Development and Impacts Management in Mountainous Areas of Northwest Yunnan, China
The Northwest Yunnan mountainous region features high biodiversity in a fragile environment. Since 1990s, with increasing amounts of visitors to the region, there is much reason for environmental concerns. However, relevant information to sustain appropriate management is lacking. Therefore, these is strong need to quantitatively describe assets and constraints of tourism development in the framework of local biodiversity protection and conservation of authentic lifestyle of local inhabitants.
Quantification of vegetation change in upper and middle Mekong River watershed using space borne remote sensing data
One of the major development issues facing the economic development of Lancang River (upper and middle Mekong River) in Yunnan Province, China, is serious environmental degradation. Environmental threats in the northern study area include a deforestation, the National Forest Protection Program (NFPP), the Slope Land Conservation Program (SLCP), mining, road construction, construction of hydropower stations, and erosion. The same threats are recognized in the southern study area. Here the expansion of rubber estates is an extra issue.
All of these issues pertain directly or indirectly to land cover or vegetation change in the two study areas. The relationship between vegetation change and human activities in these areas is a very significant socio-economic issue that is to be addressed in the research activities.
Establishment of a Centre for Eco-Management and Spatial Information Techniques, Yunnan University (P.R. of China)
This inter-universitary project is a co-operation between Belgian and Chinese institutes. The main objective is the establishment of a Centre for Eco-Management and Spatial Information Techniques (CEMSIT) at the Yunnan University (P.R. of China, Yunnan Province). The justification for the project follows from an analysis of the threefold mission of the University of Yunnan (UY), and of the Institute of Ecology and Geobotany (IEG) in particular, and its (limited) capacity to tackle scientific issues linked with the pressing problems concerning sustainable management of natural resources in the Lancang (Mekong) river catchment.
Capacity building for sustainable development in North Vietnam: Establishment and application of a geographical information system (GIS) including spatial remote sensing applications, to support sustainable environmental planning in the northern coastal provinces of Quang Ninh and Hai Phong
Vietnam faces a historical political and economical transition period, which entails development of infrastructure and industrial/economical poles; therefore, environmental management and planning are of utmost importance. Sustainable environmental management and planning require the elaboration and analysis of complex, objective and holistic data sets; Geographical Information Systems (GIS) and satellite remote sensing are powerful tools initially dedicated to achieve such objectives. The elaboration of a GIS for such purposes first demands the making of a solid environmental database. Later on, this database can be used for environmental analysis and for the preparation of documents that can support the policy making process. Since the setup of such a database is a difficult and time-consuming process, the most endangered areas should be the first to be studied. Based on this considerations, the Hai Phong / Quan Ninh region is chosen as the area of interest.
Remote sensing in evaluating the environmental impact of rangeland management
Knowledge of the processes and parameters involved in vegetation dynamics is an important aspect when land/vegetation management strategies are being designed. This is of special interest for private livestock companies and governmental organisations, which deal with an extensive form of livestock herding. This kind of herding practice typically has following main characteristic: vast land areas are used, which are usually of marginal agricultural importance for crop production such as savannahs, bush vegetation or dry mountainous areas. Because of the clear rainfall deficit compared to the potential evaporation and the lower soil fertility, these land areas have a higher ecological fragility, making a balanced vegetation-resources-management economically more important. To evaluate herding practices regarding sustainability, livestock managers are interested in quantifying the impact of these practices on the environment.
OBCD: Object-based change detection, tracking landscape dynamics with objects
Characterizations of land cover dynamics are among the most important applications of Earth Observation data,providing insights that inform management, policy and science. In this project we abandon the per-pixel paradigm and apply change detection with object-based image analysis.
In spite of its promises, the development of object-based change detection monitoring techniques still faces numerous methodological challenges. As yet, the most pressing questions are: (1) how to quantify changes between image objects, and (2) how to evaluate the accuracy of object-based change results? This project will answer these questions by evolving innovative object-based metrics for change detection, and by developing new measures to assess the accuracy of objects and their change status. Algorithmic developments will be initiated using change detection benchmark data (e.g. video sequences). Subsequently, the algorithms will be applied on multi-temporal real-world image sets covering both natural and anthropogenic landscapes, and acquired by different sensors, whereby ESA EO data from the Sentinels are particularly envisaged.
MASC: Modelling and Assessing Surface Change impacts on Belgian and Western European climate
The interactions between land surface and climate are complex. Climate changes can affect ecosystem structure and functions, by altering photosynthesis and productivity or inducing thermal and hydric stresses on plant species. These changes then impact socio-economic systems, through e.g. lowering farming or forestry incomes. Ultimately, it can lead to permanent changes in land use structure, especially when associated with other non-climatic factors, such as urbanization pressure. These interactions and changes have consequences for the climate systems, in terms of changing (1) surface properties (albedo, roughness, evapotranspiration, etc.) and (2) greenhouse gases emissions (mainly CO2, CH4, NO2). The first type of feedback alters directly the local/regional atmospheric circulation, whilst the second feedback affects the global system in the long run, through the atmospheric greenhouse gas budget. Hence this project concentrates on regional climate/land surface models that can address the first type of (short-term) feedbacks.
URBANEARS: Urban Ecosystem Analysis supported by Remote Sensing
The prospected growth in urban population and progressive global warming will put huge pressure on the quality of the environment in densely populated areas. Sustainable development and management of urban areas is hence crucial to guard the living quality in our future cities. Unfortunately, current policy support tools such as environmental models are not well adapted to the high level of heterogeneity of urban landscapes and would greatly benefit from detailed, multi-temporal, spatially distributed input data provided by remote sensing.
This project, therefore, aims at exploring the potential of the combined use of recent multi- and hyperspectral sensors in combination with structural information derived from LiDAR. Remote sensing derived information on the characteristics of green and built-up areas will be used to improve the parameterisation of urban biophysical models. As such, we strive at improving the operational value of urban ecosystem services related to temperature and water regulation.
SATHELI: Synergy of very high resolution satellite and helicopter data for the spatio-temporal characterisation of small water body dynamics
Small waterbodies exhibit considerable spatial and temporal dynamics during a single year. Flexible tools and concepts characterizing these dynamics are currently lacking. The very small scale at which small waterbodies occur makes their detection with currently available very high resolution satellite systems (spatial resolution 4 m² or more) extremely challenging. Highly innovative ground operated drones (spatial resolution 0.10 m²) will capture these habitats at very small scales and will be integrated with very high resolution satellite imagery to optimize identification and characterization of these habitats. The developed methodology will be conducted in the framework of a specific case-study on mapping the risk of liver fluke in Flanders. This parasite is transmitted by snail-hosts proliferating in the ecotone of small waterbodies. The abundance of these vectors shows seasonal peaks and is strongly influenced by water body dynamics. The goal of this case-study is to improve our spatio-temporal capacity to forecast areas under potential liver fluke threat by developing earth-observation concepts and tools to detect these small water bodies and their dynamics.
HYPERFOREST: Advanced LiDAR and airborne hyperspectral remote sensing to support forest management
Hyperspectral (HS) images have been used in a wide range of forest applications however, linking tree crowns with the spectral information they contain, such that forest diversity can be assessed, has yet to be widely demonstrated in complex forest situations. The high number of vegetation species, the occurrence of spectrally similar tree species, unpredictable spatial tree distribution, high degree of crown closure and multi-layering of the canopy; all these elements contribute to a complicated HS signal analysis. Although literature also indicates a good complementary relationship between HS and LiDAR data (as they contain very different information), few efforts exist of the integration of HS and LiDAR signals from the real data fusion perspective. On the contrary, most studies address integration by using these information sources separately in different processing phases.
WAVARS: Web-based assessment of operator performance and variability in remote sensing image analysis
Human screening and interpretation is an indispensable component in many aspects of remote sensing image analysis. Human intervention is a requisite for visual image interpretation, where the interpreter actually performs the analysis. Even in computer-based digital image processing, human screening and interpretation is still needed at certain stages. Although it is crucial for adequately assessing automated systems' performance, virtually no research has focussed on operator functioning. Instead, it is implicitly assumed that operator performance approaches perfection, and that infrequent errors are randomly distributed across time, operators and image types. The goal of the project is to test these assumptions, and to determine the human factors that influence operator functioning.
SYNOPRA: Synergy of very high resolution optical and radar data in forest mapping and inventory
Sustainable management of forest resources requires quantitative baseline data including spatial attributes and stand parameter estimates. In addition to terrestrial techniques, a number of remote sensing based methods have proved to be valuable. For many land use / land cover applications at the local and regional level the spatial resolution of the traditional high resolution satellite data, like Landsat and SPOT images, has proven to be inadequate. This is particularly true for highly fragmented forest landscapes. Scattered forest parcels fulfil diverse environmental, social, cultural and economic roles and various benefits could be obtained thought their wider integration in all land use systems (rural and urban). Very high resolution (VHR) data are also useful to identify and map planted trees in order to develop comprehensive registers, which can assist agricultural ministries with subsidy payments. In this context, the SYNOPRA project aims at developing innovative methodological techniques exploiting the synergy of optical and radar VHR data.
Evaluation of the applicability of very high resolution imagery for updating the Flemish Forest Map 2000
In Flanders, the first Flemish Forest Map (FFM 1990) has been produced from visual interpretation of colour infrared aerial photographs of the period 1978-1990 combined with terrain control operations: an expensive, labour-intensive and sometimes subjective assignment. An actualization of the FFM 1990 was conducted between 1995 and 2000 based on digital black and white orthophotos of 1995, supplemented by terrain control procedures during 1999-2000. Hence, the current FFM 2000 represents the most actual status of the forests in Flanders in terms of spatial and attribute data. In 2002, the Flemish government invested in a complete Very High Resolution (VHR) coverage of the Flemish territory. In this context this research project aims at developing an operational method to update the FFM 2000 based on the map and attribute data derived from the available IKONOS imagery.
Dynamics of spatial forest cover pattern in Flanders, based on historical and recent forest maps
The landscape is gradually gaining acceptance as a management unit. Landscape-scale management requires methods to track spatial characteristics of forest cover, and have to some extent been provided by methods developed in the field of landscape ecology. This domain of research has extensively studied forest fragmentation and has identified several consequences of forest fragmentation, such as forest area loss, decrease of forest stand size, more irregular forest boundaries and more isolated forests. Forest attributes such as size, shape, and relative position of forests in a landscape are expected to have an influence on the social, environmental and ecological functions of a forested landscape.
Sub-pixel Sharpening of High Resolution Data: Improving spatial information extraction for local and regional authorities using very high resolution data
Remotely sensed images contain a mix of pure and mixed pixels. During the process of classification, mixed pixels are commonly assigned to the class with the highest proportion of coverage to yield a hard classification. Hereby a considerable amount of information is lost. Hence soft classification has been introduced. A soft classification assigns the pixel to different classes according to the area it represents inside the pixel. However, the assignment to these classes does not specify the location inside the pixel. Atkinson suggested a method called sub-pixel mapping that divides the pixels into a certain amount of sub-pixels and allocates the different classes to the sub-pixels. The research discussed in this paper will continue from on Verhoeye’s linear approach of sub-pixel mapping. In a later stage, focus will be on application of the algorithms on high resolution (HR) imagery of urban areas (Ghent and Brussels).
