Search

Forest reserves play an important role in the provision of ecosystem goods and services; however, these reserves are facing increasing anthropogenic impacts primarily because of agricultural activities. In sub-Saharan Africa, studies on understanding the extent and impacts of agricultural expansion into forest reserves are generally lacking. Thus, this study aimed to assess the extent of agricultural expansion into forest reserves in Zambia on a national scale between 2000 and 2018 and explores the drivers. We used a remote sensing approach by employing Google Earth Engine (GEE) and Random Forests classifier to map land cover changes at five-time steps: 2000, 2005, 2010, 2015 and 2018. Landsat images were used to map six land cover classes (forest, cropland, wetland, grassland, settlement, and water body) determined by the Intergovernmental Panel on Climate Change (IPCC). A social survey was also undertaken to understand the drivers of agricultural expansion into forest reserves at two study sited located in eastern and north-western part of Zambia. Based on the land cover maps which were produced with high accuracies ranging from 82% to 94%, the results indicated a general decline of 10% in forest area and an increase of 25% in cropland. Almost 50% of the forest reserves in Zambia are experiencing some form of encroachment, and 10% of these are heavily encroached (>90% loss). Local communities indicated that population growth, loss of soil fertility, high market demand for crops, land tenure system, and lack of law enforcement are the major drivers of agricultural expansion into forest reserves. This study highlights the need for policy marker to consider trade-offs in agricultural expansion and conservation of forest reserves relation to population growth. Enhancing sustainable agricultural practices and strengthening the implementation of different legislative frameworks is key in addressing the challenges of conservation, while enhancing food security. [ABSTRACT FROM AUTHOR]

Cyclones are some of the most devastating natural disasters that cause major losses of life and property. A few months after Cyclone Idai swept through Beira, Mozambique between 4 and 20 March 2019. An analysis is presented on the impacts of this disaster by assessing the extent of the floods. This study employed Sentinel-2 imagery and remote sensing techniques including a threshold analysis and an object‑based image analysis (OBIA) approach. Water and non-water land covers were established by using a Modified Normalized Difference Water Index (MNDWI), while the classification approach assessed three land covers. These assessments were done on four dates: (1) 31 January (before), (2) 25 February 20 (before), (3) 12 March (during) and (4) 16 April (after). Accuracies ranging from 74% to 90% were attained. The OBIA approach showed that 61% of the study area was affected by floods, while the threshold approach showed that 55% of the area was flooded. Areas with low elevation and close to the coast were more effected by the floods. After the cyclone, the area under water reduced by more than 20%. The results here could be useful for developing early warning systems, risk assessment, establishing damage and developing natural hazard policies. [ABSTRACT FROM AUTHOR]

• Land cover dynamics in dry tropical environments are often overlooked. • Classification accuracy of 79–86% was attained using Landsat data. • Primary forest declined, while secondary forest increased. • Deforestation has continued to increase with low rates of recovery. • Forest connectivity declined by 22% between 1972 and 2016. Land cover dynamics in the tropical dry environments of sub-Saharan Africa are not well understood compared to humid environments, especially on a national scale. Previous studies describing land cover in the dry tropics are spatially or temporally constrained. This study presents the first long-term (1972–2016) nationwide land cover dynamics and forest connectivity analyses for Zambia. We employed 300 Landsat images and an object-based image analysis (OBIA) approach with the Random Forests (RF) classifier to map nine land covers at six time steps (1972, 1984, 1990, 2000, 2008 and 2016). Post-classification change detection was used to understand the dynamics, while landscape metrics were derived to assess the forest structural connectivity. Overall accuracies ranging from 79 to 86% were achieved. In 1972, 48% of Zambia was covered by primary forest and 16% was covered by secondary forest. By 2016, 62.74% of Zambia had undergone changes, with primary forest decreasing by 32% and secondary forest increasing by 23%. Our results showed that forests have been recovering by 0.03 to 1.3% yr−1 (53, 000–242, 000 ha yr-1); however, these rates are markedly lower than deforestation rates (- 0.54 to - 3.05% yr-1: 83, 000–453, 000 ha yr-1). Annual rates of change varied by land cover, with irrigated crops having the largest increase (+ 3.19 yr-1) and primary forest having the greatest decrease (-2.48% yr-1). Forest connectivity declined by 22%, with primary forest having the greatest decline, while the connectivity for secondary and plantation forest increased. We showed here that land cover in Zambia is highly dynamic with high rates of change, low forest recovery and decline in forest connectivity. These findings will aid in land use planning, reporting for the forthcoming 2020 Global Forest Resources Assessment, and carbon accounting, especially under the reduction in carbon emissions from deforestation and degradation programme (REDD+). [ABSTRACT FROM AUTHOR]

Highlights • Pre-processing methods were evaluated on pansharpened and standard Landsat OLI-8 using OBIA-RF. • MODTRAN atmospheric correction, DOS and cosine topographic correction were used. • A combination of MODTRAN and cosine topographic correction produced the highest accuracy of 93% on pansharpened images. • The approach is helpful in improving land cover classification accuracy when Landsat OLI-8 is used as input for OBIA-RF. Abstract The application of Landsat satellite imagery in land cover classification is affected by atmospheric and topographic errors, which have led to the development of different correction methods. In this study, moderate resolution atmospheric transmission (MODTRAN) and dark object subtraction (DOS) atmospheric corrections, and cosine topographic correction were evaluated individually and combined in a heterogeneous landscape in Zambia. These pre-processing methods were tested using a combination of object-based image analysis (OBIA) and Random Forests (RF) non-parametric classifier (hereafter referred to as OBIA-RF). This assessment aimed at understanding the combined effects of different pre-processing methods and the OBIA-RF classification method on the accuracy of Landsat operational land (OLI-8) imagery with different spatial resolutions. Here, we used pansharpened and standard Landsat OLI-8 images with 15 and 30 m spatial resolutions, respectively. The results showed that non pre-processed images reached a classification accuracy of 68% for pansharpened and 66% for standard Landsat OLI-8. Classification accuracy improved to 93% (pansharpened) and 86% (standard) when combined MODTRAN and cosine topographic correction pre-processing were applied. The results highlight the importance of pansharpening, as well as atmospheric and topographic corrections for Landsat OLI-8 imagery, when used as input in OBIA classification with the RF classifier. [ABSTRACT FROM AUTHOR]

Land cover classification of Landsat images is one of the most important applications developed from Earth observation satellites. The last four decades were marked by different developments in land cover classification methods of Landsat images. This paper reviews the developments in land cover classification methods for Landsat images from the 1970s to date and highlights key ways to optimize analysis of Landsat images in order to attain the desired results. This review suggests that the development of land cover classification methods grew alongside the launches of a new series of Landsat sensors and advancements in computer science. Most classification methods were initially developed in the 1970s and 1980s; however, many advancements in specific classifiers and algorithms have occurred in the last decade. The first methods of land cover classification to be applied to Landsat images were visual analyses in the early 1970s, followed by unsupervised and supervised pixel-based classification methods using maximum likelihood, K-means and Iterative Self-Organizing Data Analysis Technique (ISODAT) classifiers. After 1980, other methods such as sub-pixel, knowledge-based, contextual-based, object-based image analysis (OBIA) and hybrid approaches became common in land cover classification. Attaining the best classification results with Landsat images demands particular attention to the specifications of each classification method such as selecting the right training samples, choosing the appropriate segmentation scale for OBIA, pre-processing calibration, choosing the right classifier and using suitable Landsat images. All these classification methods applied on Landsat images have strengths and limitations. Most studies have reported the superior performance of OBIA on different landscapes such as agricultural areas, forests, urban settlements and wetlands; however, OBIA has challenges such as selecting the optimal segmentation scale, which can result in over or under segmentation, and the low spatial resolution of Landsat images. Other classification methods have the potential to produce accurate classification results when appropriate procedures are followed. More research is needed on the application of hybrid classifiers as they are considered more complex methods for land cover classification. [ABSTRACT FROM AUTHOR]

Taper functions are important tools for forest description, modelling, assessment, and management. A large number of studies have been conducted to develop and improve taper functions; however, few review studies have been dedicated to addressing their development and parameters. This review summarises the development of taper functions by considering their parameterisation, geographic and species-specific limitations, and applications. This study showed that there has been an increase in the number of studies of taper function and contemporary methods have been developed for the establishment of these functions. The reviewed studies also show that taper functions have been developed from simple equations in the early 1900s to complex functions in modern times. Early taper functions included polynomial, sigmoid, principal component analysis (PCA), and linear mixed functions, while contemporary machine learning (ML) approaches include artificial neural network (ANN) and random forest (RF). Further analysis of the published literature also shows that most of the studies of taper functions have been carried out in Europe and the Americas, meaning most taper equations are not specifically applicable to tropical tree species. Developing well-conditioned taper functions requires reducing the variation due to species, measurement techniques, and climatic conditions, among other factors. The information presented in this study is important for understanding and developing taper functions. Future studies can focus on developing better taper functions by incorporating emerging remote sensing and geospatial datasets, and using contemporary statistical approaches such as ANN and RF. [ABSTRACT FROM AUTHOR]

Forecasting scenarios of future intra-urban land-use (intra-urban-LU) expansion can help to curb the historically unplanned urbanization in cities in sub-Saharan Africa (SSA) and promote urban sustainability. In this study, we applied the neural network–Markov model to simulate scenarios of future intra-urban-LU expansion in Lusaka city, Zambia. Data derived from remote sensing (RS) and geographic information system (GIS) techniques including urban-LU maps (from 2000, 2005, 2010, and 2015) and selected driver variables, were used to calibrate and validate the model. We then simulated urban-LU expansion for three scenarios (business as usual/status quo, environmental conservation and protection, and strategic urban planning) to explore alternatives for attaining urban sustainability by 2030. The results revealed that Lusaka had experienced rapid urban expansion dominated by informal settlements. Scenario analysis results suggest that a business-as-usual setup is perilous, as it signals an escalating problem of unplanned settlements. The environmental conservation and protection scenario is insufficient, as most of the green spaces and forests have been depleted. The strategic urban planning scenario has the potential for attaining urban sustainability, as it predicts sufficient control of unplanned settlement expansion and protection of green spaces and forests. The study proffers guidance for strategic policy directions and creating a planning vision. [ABSTRACT FROM AUTHOR]

Forest-cover change has become an important topic in global biodiversity conservation in recent decades because of the high rates of forest loss in different parts of the world, especially in the tropical region. While human interventions are the major cause, natural disasters also contribute to forest cover changes. During the past decades, several studies have been conducted to address different aspects of forest cover changes (e.g., drivers of deforestation, degradation, interventions) in different parts of the world. In Sri Lanka, increasing rates of forest loss have been recorded during the last 100 years on a regional basis, especially in the dry zone. However, Sri Lanka needs detailed studies that employ contemporary data and robust analytical tools to understand the patterns of forest cover changes and their drivers. The dry zone of Sri Lanka encompasses 59% of the total land area of the country, ergo, the most extensive forest cover. Our study analyzed forest cover dynamics and its drivers between 1992 and 2019. Our specific objectives included (i) producing a forest cover map for 2019, (ii) analyzing the spatiotemporal patterns of forest cover changes from 1992 to 2019, and (iii) determining the main driving forces. Landsat 8 images were used to develop forest-cover maps for 2019, and the rest of the forest cover maps (1992, 1999, and 2010) were obtained from the Forest Department of Sri Lanka. In this study, we found that the dry zone had undergone rapid forest loss (246,958.4 ha) during the past 27 years, which accounts for 8.0% of the net forest cover changes. From 2010 to 2019, the rates of forest loss were high, and this can be associated with the rapid infrastructure development of the country. The findings of this study can be used as a proxy to reform current forest policies and enhance the forest sustainability of the study area. [ABSTRACT FROM AUTHOR]

The advancement in satellite remote sensing technology has revolutionised the approaches to monitoring the Earth's surface. The development of the Copernicus Programme by the European Space Agency (ESA) and the European Union (EU) has contributed to the effective monitoring of the Earth's surface by producing the Sentinel-2 multispectral products. Sentinel-2 satellites are the second constellation of the ESA Sentinel missions and carry onboard multispectral scanners. The primary objective of the Sentinel-2 mission is to provide high resolution satellite data for land cover/use monitoring, climate change and disaster monitoring, as well as complementing the other satellite missions such as Landsat. Since the launch of Sentinel-2 multispectral instruments in 2015, there have been many studies on land cover/use classification which use Sentinel-2 images. However, no review studies have been dedicated to the application of ESA Sentinel-2 land cover/use monitoring. Therefore, this review focuses on two aspects: (1) assessing the contribution of ESA Sentinel-2 to land cover/use classification, and (2) exploring the performance of Sentinel-2 data in different applications (e.g., forest, urban area and natural hazard monitoring). The present review shows that Sentinel-2 has a positive impact on land cover/use monitoring, specifically in monitoring of crop, forests, urban areas, and water resources. The contemporary high adoption and application of Sentinel-2 can be attributed to the higher spatial resolution (10 m) than other medium spatial resolution images, the high temporal resolution of 5 days and the availability of the red-edge bands with multiple applications. The ability to integrate Sentinel-2 data with other remotely sensed data, as part of data analysis, improves the overall accuracy (OA) when working with Sentinel-2 images. The free access policy drives the increasing use of Sentinel-2 data, especially in developing countries where financial resources for the acquisition of remotely sensed data are limited. The literature also shows that the use of Sentinel-2 data produces high accuracies (>80%) with machine-learning classifiers such as support vector machine (SVM) and Random forest (RF). However, other classifiers such as maximum likelihood analysis are also common. Although Sentinel-2 offers many opportunities for land cover/use classification, there are challenges which include mismatching with Landsat OLI-8 data, a lack of thermal bands, and the differences in spatial resolution among the bands of Sentinel-2. Sentinel-2 data show promise and have the potential to contribute significantly towards land cover/use monitoring. [ABSTRACT FROM AUTHOR]

Decision tree (DT) algorithms are important non-parametric tools used for land cover classification. While different DTs have been applied to Landsat land cover classification, their individual classification accuracies and performance have not been compared, especially on their effectiveness to produce accurate thresholds for developing rulesets for object-based land cover classification. Here, the focus was on comparing the performance of five DT algorithms: Tree, C5.0, Rpart, Ipred, and Party. These DT algorithms were used to classify ten land cover classes using Landsat 8 images on the Copperbelt Province of Zambia. Classification was done using object-based image analysis (OBIA) through the development of rulesets with thresholds defined by the DTs. The performance of the DT algorithms was assessed based on: (1) DT accuracy through cross-validation; (2) land cover classification accuracy of thematic maps; and (3) other structure properties such as the sizes of the tree diagrams and variable selection abilities. The results indicate that only the rulesets developed from DT algorithms with simple structures and a minimum number of variables produced high land cover classification accuracies (overall accuracy > 88%). Thus, algorithms such as Tree and Rpart produced higher classification results as compared to C5.0 and Party DT algorithms, which involve many variables in classification. This high accuracy has been attributed to the ability to minimize overfitting and the capacity to handle noise in the data during training by the Tree and Rpart DTs. The study produced new insights on the formal selection of DT algorithms for OBIA ruleset development. Therefore, the Tree and Rpart algorithms could be used for developing rulesets because they produce high land cover classification accuracies and have simple structures. As an avenue of future studies, the performance of DT algorithms can be compared with contemporary machine-learning classifiers (e.g., Random Forest and Support Vector Machine). [ABSTRACT FROM AUTHOR]

Land cover change (LCC) has significant effects on the global ecosystem diversity and function. This topic has received increasing attention due, in part, to its relationship with climate change, and the availability of remotely-sensed imagery that is used to monitor LCC. However, studies analysing the factors that drive LCC at large spatial scales and over long temporal scales are uncommon. This study aimed to identify the factors driving long-term (44 years, 1972–2016) national level LCC in Zambia. Two analyses were conducted, with the first considering factors that led to any LCC. The second scenario identified factors associated with changes from forests to other land covers, and the reversion to forests from non-forested covers. Candidate factors considered in both analyses include accessibility, proximity, topography, climate, conservation and socioeconomics. A classification tree (CT) approach was used to relate the explanatory candidate factors to LCC. The results showed that the CT models predicted LCC with accuracies ranging from 71 to 85%. The first analysis showed that the major factors determining LCC were percentage of area under agriculture, distance to water bodies, change in crop yield, mean temperature and elevation. Meanwhile, the second analysis showed that primary, secondary and plantation forest cover losses were mainly influenced by human population density, crop yield per hectare and mean crop yield, respectively. Protection status was the most important factor for forest reversion and recovery, while a variety of factors including distance to the railway, elevation and total precipitation also influenced forest reversion and recovery. The findings from this study provide insights into the factors that influence LCC and are important for developing effective policies to reduce the negative impacts of LCC and to promote forest reversion and recovery through effective management of protected areas. While this study focused on factors associated with historical LCC, the findings will also help to predict and understand future LCC scenarios. Unlabelled Image • Long-term (1972–2016) land cover change is influenced by different factors in Zambia. • Factors of land cover change were assessed using the Classification Tree analysis. • The CT models had accuracies ranging from 71 to 85%. • Forest losses were influenced by crop yield, cultivated area, population density and distance to facilities (e.g. roads, railway and towns). • Protected status (e.g. national forests) was important for forest recovery and reversion. [ABSTRACT FROM AUTHOR]