Article

Mapping potential, existing and efficient wetlands using free remote sensing data

Authors:
  • CNRS, Rennes, France
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

Although wetlands remain threatened by human pressures and climate change, monitoring and managing them are challenging due to their high spatial and temporal dynamics within a fine-grained pattern. New satellite time-series at high temporal and spatial resolutions provide a promising opportunity to map and monitor wetlands. The objective of this study was to develop an operational method for managing valley bottom wetlands based on available free remote sensing data. The Potential, Existing, Efficient Wetlands (PEEW) approach was adapted to remote sensing data to delineate three wetland components: (1) potential wetlands, mapped from a digital terrain model derived from LiDAR data; (2) existing wetlands, delineated from land cover maps derived from Sentinel-1/2 time-series; and (3) efficient wetlands, identified from functional indicators (i.e. annual primary production, vegetation phenology, seasonality of carbon flux) derived from MODIS annual time-series. Soil and vegetation samples were collected in the field to calibrate and validate classification of remote sensing data. The method was applied to a 113 000 ha watershed in northwestern France. Results show that potential wetlands were successfully delineated (82% overall accuracy) and covered 21% of the watershed area, while 44% of existing wetlands had been lost. Small wetlands along headwater channels, which are considered as ordinary, cover 56% of wetland area in the watershed. Efficient wetlands were identified as contiguous pixels with a similar temporal functional trajectory. This method, based on free remote sensing data, provides a new perspective for wetland management. The method can identify sites where restoration measures should be prioritized and enables better understanding and monitoring of the influence of management practices and climate on wetland functions.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... The accurate identification of physical settings comprising of peat is pertinent to digital soil mapping and a variety of associated land management activities. These applications can conform to the ascertainment of areas with higher soil carbon and organic matter contents (Vitt et al., 2009), as well as detecting wetlands (Rapinel et al., 2019;Whitcomb et al., 2009) and evaluating the extent of environmental impact for these respective regions (Sulla-Menashe et al., 2018). The recent availability and enhanced resolution of remotely-sensed imagery from multiple platforms, has facilitated digital soil mapping research for peatlands (Minasny et al., 2019). ...
Article
Full-text available
Neural networks were explored to achieve a binary classification for determining land corresponding to peat for a study area in the boreal forest of northern Ontario, Canada. Environmental covariates were employed as predictors and obtained from multiple sources, which included multispectral imagery, LiDAR, SAR, and aeromagnetic data. A dense neural network (DNN), as well as a convolutional neural network (CNN), were each implemented. Logistic regression, support vector machine (SVM) and random forest (RF) approaches were also modelled. Neighboring pixels surrounding the soil sampling sites were incorporated as input into the CNN, that permitted training on additional information that was not exploited by other methods. Preliminary results indicate that a CNN can attain improved accuracies for peat classification, when compared against other approaches.
... Values ranging from 5 to 50, up in fives, were assessed for both parameters. Gini index was used as a tree node splitting criterion (Rapinel et al., 2019). The classification was performed by the random forest algorithm in ArcGIS Pro 2.5 (Learn ArcGIS) software. ...
Article
Full-text available
Resilience is the ability of a system to absorb disturbances, rearrange itself, and adapt in order to maintain its functionality, structure, identity, and feedback. Research involving fire resilience in subtropical wetlands (SW) allows us to understand the dynamics of these ecosystems, measure impacts on fauna and flora, and promote policies for the management and protection. The aim of the present study is to assess the fire resilience of SW. The study was divided into three steps: (i) burned area classification, (ii) vegetation pattern classification, and (iii) temporal analysis of SW fire resilience based on NDVI calculation. Our results show that (a) high resilience potential of emerging plants, which developed green leaves in less than 90 days after the fire; (b) poor recovery of peatlands with underground fire history. Daily coverage of high spatial resolution PlanetScope images has great potential for classification and monitoring of land use in areas where there are rapid changes, such as after a fire event, explosions, and dam ruptures with ore tailings, for example.
... Multi-source remote sensing data including optical, hyperspectral, and SAR images have been used to identify worldwide wetlands (Guo et al., 2017;Rapinel et al., 2019;Wohlfart et al., 2018). Landsat images have been widely used due to the broad and long term coverage, fine resolution, and free availability (Amani et al., 2019b, Mao et al., 2019b. ...
Article
Spatially and thematically explicit information of wetlands is important to understanding ecosystem functions and services, as well as for establishment of management policy and implementation. However, accurate wetland mapping is limited due to lacking an operational classification system and an effective classification approach at a large scale. This study was aimed to map wetlands in China by developing a hybrid object-based and hierarchical classification approach (HOHC) and a new wetland classification system for remote sensing. Application of the hybrid approach and the wetland classification system to Landsat 8 Operational Land Imager data resulted in a wetland map of China with an overall classification accuracy of 95.1%. This national scale wetland map, so named CAS_Wetlands, reveals that China’s wetland area is estimated to be 451,084 ± 2014 km², of which 70.5% is accounted by inland wetlands. Of the 14 sub-categories, inland marsh has the largest area (152,429 ± 373 km²), while coastal swamp has the smallest coverage (259 ± 15 km²). Geospatial variations in wetland areas at multiple scales indicate that China’s wetlands mostly present in Tibet, Qinghai, Inner Mongolia, Heilongjiang, and Xinjiang Provinces. This new map provides a new baseline data to establish multi-temporal and continuous datasets for China’s wetlands and biodiversity conservation.
... Potential wetlands mapping allows the delineation of areas where water-related ecosystems could occur with a high level of probability. The delineation of (potential) wetlands often relies on topographic indices or their combination with EO data (Beven and Kirkby, 1979;Bwangoy et al., 2010;Agren et al. 2014;Ludwig et al., 2019;Rapinel et al., 2019). The present assessment was based on the mapping of open surface water dynamics (SWD) using optical Landsat-8 time series for the year 2015, which was combined with hydrological and topographic indices derived from a DEM (EU-DEM with~25 m of spatial resolution). ...
Article
Wetlands are the most fragile and threatened ecosystems worldwide, and also one of the most rapidly declining. At the same time wetlands are typically biodiversity hotspots and provide a range of valuable ecosystem services, such as water supply and purification, disaster risk reduction, climate change adaptation, and carbon sequestration. Pressures on wetlands are likely to further intensify in the coming decades due to increased global demand for land and water, and due to climate change. Stakeholders at all levels of governance have to be involved to slow, stop and reverse these processes. However, the information they need on wetland extent, their ecological character, and their ecosystem services is often scattered, sparse and difficult to find and access. The freely available Sentinel satellite data of the Copernicus Programme, as well as the Landsat archive, provide a comprehensive basis to map and inventory wetland areas (extent), to derive information on the ecological status, as well as long- and short-term trends in wetland characteristics. However, making use of these Earth Observation (EO) resources for robust and standardized wetland monitoring requires expert knowledge on often complex data processing techniques, which impedes practical implementation. In this respect, the Satellite-based Wetland Observation Service (SWOS), a Horizon 2020 funded project (www.swos-service.eu) has developed and made disseminated monitoring approaches based on EO data, specifically designed for less experienced satellite data users. The SWOS monitoring tools aim at assisting countries in conducting national wetland inventories for their Sustainable Development Goals (SDG) reporting and monitoring obligations, and additionally facilitates other monitoring obligations such as those required by the Ramsar Convention and supports decision-making in local conservation activities. The four main components of the SWOS approach are: map and indicator production; software development; capacity building; and initializing the GEO Wetlands Community Portal. Wetland managers and data analysists from more than fifty wetland sites and river basins across Europe, the Middle East, and Africa investigated the benefits and limitations of this EO-based wetland mapping and monitoring approach. We describe research that applies the SWOS tools to test their potential for the mapping of wetlands in a case study based in Albania, and show its effectiveness to derive metrics relevant to the monitoring of SDG indicator 6.6.1.
... Accurate wetland vegetation mapping is needed to distinguish the different classes of macrophyte vegetation, understand their functions, and analyze the response of wetlands to climate change and human activities. Recently, remote sensing methodology for the study of wetland vegetation has been used by many researchers [4][5][6][7][8][9][10][11]. ...
Article
Full-text available
In wetland environments, vegetation has an important role in ecological functioning. The main goal of this work is to identify an optimal combination of Sentinel-1 (S1), Sentinel-2 (S2), and Pleiades data using ground-reference data to accurately map wetland macrophytes in the Danube Delta. We tested several combinations of optical and Synthetic Aperture Radar (SAR) data rigorously at two levels. First, in order to reduce the confusion between reed (Phragmites australis (Cav.) Trin. ex Steud.) and other macrophyte communities, a time series analysis of S1 data was performed. The potential of S1 for detection of compact reed on plaur, compact reed on plaur/reed cut, open reed on plaur, pure reed, and reed on salinized soil was evaluated through time series of backscatter coefficient and coherence ratio images, calculated mainly according to the phenology of the reed. The analysis of backscattering coefficients allowed separation of reed classes that strongly overlapped. The coherence coefficient showed that C-band SAR repeat pass interferometric coherence for cut reed detection is feasible. In the second section, random forest (RF) classification was applied to the S2, Pleiades, and S1 data and in situ observations to discriminate and map reed against other aquatic macrophytes (submerged aquatic vegetation (SAV), emergent macrophytes, some floating broad-leaved and floating vegetation of delta lakes). In addition, different optical indices were included in the RF. A total of 67 classification models were made in several sensor combinations with two series of validation samples (with the reed and without reed) using both a simple and more detailed classification schema. The results showed that reed is completely discriminable compared to other macrophytes communities with all sensor combinations. In all combinations, the model-based producer's accuracy (PA) and user's accuracy (UA) for reed with both nomenclatures were over 90%. The diverse combinations of sensors were valuable for improving the overall classification accuracy of all of the communities of aquatic macrophytes except Myriophyllum spicatum L.
... The model supported interrelationships between soil, vegetation, and remotely sensed information and established reasonable predictions for soil organic carbon and bulk density in a coastal wetland. On the other hand, Potential, Existing, Efficient Wetland (PEEW) approach as an operational method was developed and utilized to delineate three wetland components using remote-sensing data (Rapinel et al., 2019). The potential wetlands derived from light detection and ranging (LiDAR) data, existing wetlands from Sentinel-1/2 times series, and existing wetlands from Moderate Resolution Imaging Spectroradiometer (MODIS) annual time series provided a new perspective for wetland management by prioritizing the sites for restoration measures. ...
Article
Full-text available
This article presents an update on the research and practical demonstration of wetland‐based treatment technologies for protecting water resources and environment covering papers published in 2019. Wetland applications in wastewater treatment, stormwater management, and removal of nutrients, metals and emerging pollutants including pathogens are highlighted. A summary of studies focusing on the effects of vegetation, wetland design and operation strategies, and process configurations and modeling, for efficient treatment of various municipal and industrial wastewaters is included. In addition, hybrid and innovative processes with wetlands as a platform treatment technology are presented.
... Accurate wetland vegetation mapping is needed to distinguish the different classes of macrophyte vegetation, understand their functions, and analyze the response of wetlands to climate change and human activities. Recently, remote sensing methodology for the study of wetland vegetation has been used by many researchers [4][5][6][7][8][9][10][11]. ...
... Remote sensing is a potentially very useful source of information for spatial monitoring of natural or cultivated vegetation. Many approaches have been developed recently for monitoring and mapping of vegetation (Roberts et al., 2015;Niculescu et al., 2016 ;Niculescu et al., 2017 ;Shivers et al., 2018;Muller-Karger et al., 2018;Rapinel et al., 2019). We noted the new opportunities for the monitoring of the vegetation with the arrival the European multi-sensor images in time series and namely with the increasing number of satellites and the availability of free data. ...
Article
Full-text available
Remote sensing is a potentially very useful source of information for spatial monitoring of natural or cultivated vegetation. The latest advances, in particular the arrival of new image acquisition programs, are changing the temporal approach to monitoring vegetation. The latest European satellites launched, delivering an image every 5 days for each point on the globe, allow the end of a growing season to be monitored. The main objective of this work is to identify and map the vegetation in the Pays de Brest area by using a multi sensors stacking of Sentinel-1 and Sentinel-2 satellites data via Random Forest, Rotation forests (RoF) and Canonical Correlation Forests (CCFs). RoF and CCF create diverse base learners using data transformation and subset features. Twenty four radar images and optical dataa representing different dates in 2017 were processed in time series stacks. The results of RoF and CCF were compared with the ones of RF.
... Estimated to cover 58.2 million hectares (Prasad et al. 2002), wetlands in India are facing tremendous pressures owing to reckless urbanization (Anand and Oinam 2020;Roy-Basu et al. 2020), deforestation, and land system changes (Pandit et al. 2007;Reddy et al. 2016a) that have modified their trophic status (Vass et al. 2015). Over the last two decades, earth observation data from remotely sensed platforms has been widely used to quantify land system changes around wetlands both globally (Ozesmi and Bauer 2002;Rapinel et al. 2019;Kandus et al. 2018) and within India (Selvam et al. 2003;Shanmugam et al. 2006;Reddy et al. 2016b). Such information integrated with field-based observations and input into geospatial models has provided new insights into understanding the wetland dynamics (Chatterjee et al. 2015;Mondal et al. 2017;Saha and Pal 2019). ...
Article
Full-text available
Wetlands are among the most vulnerable and dynamic ecosystems of the world. Any change in the anthropogenic footprint or climate affects the health of these pristine ecologically and socioeconomically important ecosystems. In the present study, land use land cover changes (LULCC) and fragmentation of natural landscape changes in an urban wetland, Khushalsar, located in the heart of the Srinagar City, were assessed using high-resolution satellite data, geospatial modeling approach, and ground observations over the last ~ 5 decades (1965 and 2018). The spatiotemporal changes in LULC of the wetland were assessed for 3 time periods that include 1965–1980, 1980–2018, and 1965–2018. Additionally, landscape fragmentation tool (LFT) was used to quantify fragmentation of land cover. The analysis of LULCC indicated that built-up areas in the vicinity of the wetland increased by 510% between 1965 and 2018. The aquatic vegetation and marshy lands increased by 150% and 33% respectively. The area under agriculture, plantation, open water, and barren lands decreased mostly taken over by built-up areas. Within the wetland, the area under open water spread reduced by 75% while the aquatic vegetation increased by 150% from 1965 to 2018. The built-up areas including roads also showed a substantial increase. The LFT analysis revealed four categories of landscapes i.e., patch, edge, perforated, and core areas. Since the natural land cover types were taken over by land use predominantly built-up areas, the core natural areas and perforated landscapes in the Khushalsar shrunk by 34% and 94% respectively indicating fragmentation of natural environment. The lack of sewage treatment facility, reckless unplanned urbanization within and in the vicinity of the wetland, is responsible for the degradation of the Khushalsar wetland.
... The results showed that the condition of the wetland has deteriorated over the past three decades. Rapinel, et al. [38] mapped and monitored wetlands condition on the Couesnon watershed, located in northeastern Brittany, France, using LiDAR, Sentinel-1/2 and Moderate Resolution Imaging Spectroradiometer (MODIS) time-series data. The results showed that 44 % of the existing wetlands have been destroyed. ...
Article
Full-text available
Due to the excessive use of natural resources in the contemporary world, the importance of ecological and environmental condition modeling has increased. Wetlands and cities represent the natural and artificial strategic areas that affect ecosystem conditions. Changes in the ecological conditions of these areas have a great impact on the conditions of the global ecosystem. Therefore, modeling spatiotemporal variations of the ecological conditions in these areas is critical. This study was aimed at comparing degrees of variation among surface ecological conditions due to natural and unnatural factors. Consequently, the surface ecological conditions of Gomishan city and Gomishan wetland in Iran were modeled for a period of 30 years, and the spatiotemporal variations were evaluated and compared with each other. To this end, 20 Landsat 5, 7, and 8, and 432 Moderate Resolution Imaging Spectroradiometer (MODIS), monthly land surface temperature (LST) (MOD11C3) and normalized difference vegetation index (NDVI) (MOD13C3) products were utilized. The surface ecological conditions were modeled according to the Remote Sensing-based Ecological Index (RSEI), and the spatiotemporal variation of the RSEI values in the study area (Gomishan city, Gomishan wetland) were evaluated and compared with each other. According to MODIS products, the mean of the LST and NDVI variance values for the study area (Gomishan city, Gomishan wetland) were obtained to be 6.5 °C (2.1, 12.1) and 0.009 (0.005, 0.013), respectively. The highest LST and NDVI temporal variations were found for Gomishan wetland near the Caspian Sea. According to Landsat images, Gomishan wetland and Gomishan city have the highest and lowest temporal variations in surface biophysical characteristics, respectively. The mean RSEI for the study area (Gomishan city, Gomishan wetland) was 0.43 (0.65, 0.29), respectively. Additionally, the mean Coefficient of Variation (CV) of RSEI for the study area (Gomishan city, Gomishan wetland) was 0.10 (0.88, 0.51), respectively. The surface ecological conditions of Gomishan city were worse than those of the Gomishan wetland at all dates. Temporal variations in the surface ecological conditions of Gomishan wetland were greater than those of the study area and Gomishan city. These results can provide useful and effective information for environmental planning and decision-making to improve ecological conditions, protect the environment, and support sustainable ecosystem development.
... LiDAR retrieval data can be analyzed to generate environmental covariates corresponding to vegetation. To our knowledge, for analysis over study areas greater than a few ten square kilometers, LiDAR has only been applied in DSM for generating finer spatial resolution DEMs (Akumu et al., 2015;Anderson et al., 2006;Greve et al., 2012;Mulder et al., 2011), and in particular, recently for wetland environments (O'Neil et al., 2019;Rapinel et al., 2019;Vernimmen et al., 2019). Environmental covariates for parent material, if specified, tend to be limited to vector files corresponding to surficial geology (Minasny et al., 2019) or mineralogy (Mulder et al., 2011). ...
Article
This study aimed to improve the accuracy in modelling the properties of forest soil by employing a variety of remotely sensed data. In addition to the commonly used multispectral satellite imagery, airborne LiDAR (light detection and ranging) data were exploited to derive detailed vegetation properties and topographic variables. Random forest (RF) and support vector machine (SVM) approaches were applied to classify soil types in terms of texture, calcareous substrate reaction to hydrochloric acid, and ELC (Ecological Land Classification) moisture regime. The developed methods were tested on data acquired over a boreal region (49° - 50° N, 81° - 84° W) with a combined area of 4,085 km² in the Great Clay Belt (GCB) region, Ontario, Canada. Compared with the field-collected data, the overall accuracies and kappa coefficients of the retrieved soil properties were greater than 0.7 and 0.5, respectively. The accuracies attained between the RF and SVM approaches were similar, but in general the highest accuracies were achieved by the RF method. The models developed for the whole GCB regions generated accuracies comparable to those for the three sub-regions. The lowest modelling uncertainties occurred in areas dominated by peatland, whereas the highest modelling uncertainties existed in the regions with dry moisture regime or clayey soil at the surface. The results also showed that environmental covariates corresponding to vegetation were most important in the prediction of soil properties. Specifically, canopy height model (CHM) and gap fraction derived from LiDAR data, were among the most important variables. The inclusion of LiDAR-derived covariates demonstrated potential, applied in addition with topographic and climatic covariates and optical imagery. CHM pertains to the vertical dimension, and gap fraction relates to the density of the canopy layer, respectively; both covariates that offer supplemental detail that is not necessarily ascertained for the canopy by optical imagery.
... These wetlands merge with the surrounding terrestrial ecosystem [1] posing challenges when mapping their spatial extent using optical sensors, especially during the dry period when the surrounding and wetland vegetation are not very healthy, resulting in similar spectral reflectance of soils and other land-cover classes. Rapinel et al., [37] reported that inventorying and characterization of wetlands in semi-arid and arid areas are limited to mostly small basins. Furthermore, Cape et al., [38] reported that the semi-arid wetlands are lost over a short period of time because of the anthropogenic activities including overexploitation of their water for irrigation. ...
Article
Full-text available
Wetlands are ranked as very diverse ecosystems, covering about 4-6% of the global land surface. They occupy the transition zones between aquatic and terrestrial environments, and share characteristics of both zones. Wetlands play critical roles in the hydrological cycle, sustaining livelihoods and aquatic life, and biodiversity. Poor management of wetlands results in the loss of critical ecosystems goods and services. Globally, wetlands are degrading at a fast rate due to global environmental change and anthropogenic activities. This requires holistic monitoring, assessment, and management of wetlands to prevent further degradation and losses. Remote-sensing data offer an opportunity to assess changes in the status of wetlands including their spatial coverage. So far, a number of studies have been conducted using remotely sensed data to assess and monitor wetland status in semi-arid and arid regions. A literature search shows a significant increase in the number of papers published during the 2000-2020 period, with most of these studies being in semi-arid regions in Australia and China, and few in the sub-Saharan Africa. This paper reviews progress made in the use of remote sensing in detecting and monitoring of the semi-arid and arid wetlands, and focuses particularly on new insights in detection and monitoring of wetlands using freely available multispectral sensors. The paper firstly describes important characteristics of wetlands in semi-arid and arid regions that require monitoring in order to improve their management. Secondly, the use of freely available multispectral imagery for compiling wetland inventories is reviewed. Thirdly, the challenges of using freely available multispectral imagery in mapping and monitoring wetlands dynamics like inundation, vegetation cover and extent, are examined. Lastly, algorithms for image classification as well as challenges associated with their uses and possible future research are summarised. However, there are concerns regarding whether the spatial and temporal resolutions of some of the remote-sensing data enable accurate monitoring of wetlands of varying sizes. Furthermore, it was noted that there were challenges associated with the both spatial and spectral resolutions of data used when mapping and monitoring wetlands. However, advancements in remote-sensing and data analytics provides new opportunities for further research on wetland monitoring and assessment across various scales.
... Due to the non-existence of superficial connection with the river, the functionality between soil-water and aquifer in the geographically isolated wetlands is influenced by the structure of hydric soils and internal topography, both determining the lateral and subsurface flows, controlling downstream flow and acting as discharge and recharge areas [58]. Figure 3 shows the internal pattern of the studied wetland: (1) shrub and tall grass vegetation that dominate in the border of wetland and (2) short-grass vegetation organized in circular patches located exclusively in the Each pattern represents hydrologic soils influenced by distinct microtopography, vegetation, and local distribution of water [59][60][61]. In the Cerrado biome, vegetation structure is strongly related to groundwater depth [19]. ...
Article
Full-text available
In seasonal flooding isolated wetlands, the degree of wetness suggests a close synergy between soil processes, landscape evolution and hydrology along space and time. Until now, that subject has received insufficient attention despite natural wetlands supply essential environmental services to society and are surrounded by intensive agriculture that uses agro-chemicals and fertilizers in their management. The objectives of this study were to propose an infiltration architecture model based on local surface and subsurface water-fluxes in isolated wetland embedded in lateritic plateau covered by savanna and qualify the environmental sensitivity as an area of aquifer recharge. Grain size, soil bulk density, and hydraulic conductivity were determined in five profiles in a soil catena. Unmanned Aerial Vehicle high-resolution images were obtained to generate a digital elevation model and discriminate areas with different vegetation, water accumulation, and environmental sensitivity. Electrical tomography was performed to unveil the soil architecture and infiltration. The soils (Plinthosols) developed on aquic conditions determine the linkage between the surface-subsurface hydrodynam-ics with the soil's physical properties. We have identified vertical and lateral water-flows in the soil architecture. Vertical flow occurs exclusively at the center, where the wetland is characterized as a recharge zone. Lateral flow towards the borders characterizes a discharge zone. The recharge zone is a depression surrounded by crops; therefore, it is a point of high environmental sensitivity. This hydrodynamic model is essential to support studies related to the dispersion of contaminants since soybean agriculture dominates the whole area of well-drained soils in the Brazilian Cerrado.
... For both parameters, values between 5 and 50 were assessed, varying by 5. The Gini index was used as the tree node division criterion [44]. k-NN is based on the determination of training samples that are closer to unclassified data, classifying them in relation to the most frequent class of the selected neighbors [30]. ...
Article
Full-text available
The classification of vegetation species is a fundamental technical task, necessary for the sustainable management of wetland ecosystems. GEographic-Object-Based Image Analysis (GEOBIA) and data mining have enabled classification and monitoring of wetlands with higher accuracies and lower costs. The objective of this study is to evaluate the performance of Random Forest (RF) and k-Nearest Neighbor (k-NN) data mining methods in vegetation species object-based classification in a subtropical wetland, integrating Sentinel-1 and Sentinel-2A images. In this work, 91.3% accuracy was reached for the object-based classification of vegetation in wetlands with the RF method, and 81.1% accuracy was reached with the k-NN method. Synthetic aperture radar (SAR) features obtained the two major importances, 9.7% (VHM) and 8.9% (VVM). The optical features, red edge and the two short-wavelength infrared bands resulted in values greater than 6%. We conclude that the integration of optical satellite images and SAR, together with the use of GEOBIA and data mining, was successful in classifying vegetation classes of wetlands.
... They have many unique functions such as erosion control, coastal protection, flood prevention, regulation of water regime, biodiversity conservation, maintain and sustain of wildlife, water quality improvement, control of climate change, and use for ecotourism and recreation (Bwangoy et al. 2010;Chen et al. 2014a;Debanshi and Pal 2020). On the other hand, mistaken agricultural policies, industrialization, urbanization, and inconvenient land use decisions seriously harm those environments, and so, total areas of lakes are decreasing (Rapinel et al. 2019). Hence, it is an exceedingly considerable issue to monitor the lakes and extract their surface areas in order to protect natural heritage zones and transfer them to posterity. ...
Article
Full-text available
The surface areas of lakes alter constantly due to many factors such as climate change, land use policies, and human interventions, and their surface areas tend to decrease. It is necessary for obtain baseline datasets such as surface areas and boundaries of water bodies with high accuracy, effectively, economically, and practically by using satellite images in terms of management and planning of lakes. Extracting surface areas of water bodies using image classification algorithms and high-resolution RGB satellite images and evaluating the effectiveness of different image classification algorithms have become an important research domain. In this experimental study, eight different machine learning-based classification approaches, namely, k-nearest neighborhood (kNN), subspaced kNN, support vector machines (SVMs), random forest (RF), bagged tree (BT), Naive Bayes (NB), and linear discriminant (LD), have been utilized to extract the surface areas of lakes. Lastly, autoencoder (AE) classification algorithm was applied, and the effectiveness of all those algorithms was compared. Experimental studies were carried out on three different lakes (Hazar Lake, Salda Lake, Manyas Lake) using high-resolution Turkish RASAT RGB satellite images. The results indicated that AE algorithm obtained the highest accuracy values in both quantitative and qualitative analyses. Another important aspect of this study is that Structural Similarity Index (SSIM) and Universal Image Quality Index (UIQI) metrics that can evaluate close to human perception are used for comparison. With this application, it has been shown that overall accuracy calculated from test data may be inadequate in some cases by using SSIM, UIQI, mean squared error (MSE), peak signal to noise ratio (PSNR), and Cohen’s KAPPA metrics. In the last application, the robustness of AE was examined with boxplots.
... However, due to the remote and difficult access of most wetlands, the traditional data collection method is labor-intensive, costly, and dangerous (Mahdianpari et al., 2018a(Mahdianpari et al., , 2018bMohammadimanesh et al., 2018). Remote sensing technology allows for large-scale synchronous observation at a high time resolution, which solves the limitations of traditional technology and has become the most effective means to obtain spatiotemporal information on wetlands (Taddeo et al., 2019;Abeysinghe et al., 2019;Rapinel et al., 2019;Lane et al., 2014;Franklin et al., 2017). ...
Article
Full-text available
The accurate classification of wetland vegetation is essential for rapid assessment and management. The Honghe National Nature Reserve (HNNR), located in Northeast China, was studied. The multi-scale remote sensing data of a new generation of Chinese high-spatial-resolution earth observation satellites Gaofen-1 (GF-1), Gaofen-2 (GF-2), Ziyuan-3 (ZY-3), and international earth observation satellites Sentinel-2A and Landsat 8 OLI were selected as sources. Based on the DeepLabV3 Plus deep learning model, 12 intelligent marsh vegetation classification models were constructed. We quantitatively analyzed the applicability and identification ability of DeepLabV3 Plus for classifying complex marsh vegetation. We discuss the differences in accuracy of marsh vegetation classification with different remote sensing data sets. The spatial resolution of remote sensing data sets ranges from 30 m to 0.8 m, and spectral bands range from blue bands (450 nm) to shortwave infrared bands (2280 nm). The specific conclusions of this study are as follows: (1) The DeepLabV3 Plus model better identified marsh vegetation, but there were significant differences in the classification accuracy of 12 DeepLabV3 Plus intelligent identification models. (2) Under the same conditions of the spectral bands (four Blue ~ NIR bands), the accuracy of deep-water marsh vegetation classification gradually increased as spatial resolution improved. For shallow-water marsh vegetation, when the accuracy of vegetation classification increased to a certain level, the classification accuracy decreased with the improvement of spatial resolution, which indicated that high-resolution images reduced pixel mixing to a certain extent, but for some vegetation types, the internal spectral difference increased, which made classification more difficult. (3) The increase of spectral bands improved the classification of marsh vegetation, while the classification accuracy of models with spectral indices was better than that of models only including spectral bands. (4) The accuracy of marsh vegetation classification was greatly improved by combining spectral bands and spectral indices. (5) The classification of the five sensor satellite images had statistical differences between models with different spatial resolutions and models with different spectral ranges.
... In recent years, remote sensing technology has become a vital tool for large-scale wetland research because of its high efficiency, wide spatial coverage, and small restrictions caused by topographical conditions McCarthy et al., 2018). Currently, the remote sensing data used for wetland identification and classification mainly include optical images, synthetic aperture radar (SAR) images, LiDAR images, and multi-source remote sensing data (Rapinel et al., 2019;Kumar et al., 2013). Because of the rich spectral information and diverse spatial resolution from low (>1 km) to high (<1 m), optical remote sensing data have been widely used in wetland classification studies carried out at different scales (Mahdianpari et al., 2018). ...
Article
Full-text available
Wetland vegetation is susceptible to climate change and human disturbance, and has experienced significant losses and degradation. However, the spatial patterns of dynamics for China’s inland lake wetlands remain unknown. In this paper, an adaptive-stacking algorithm based on Google Earth Engine was proposed to map the vegetation distribution of Dongting Lake wetland using Sentinel-1/2 and DEM data. Subsequently, LandTrendr was utilized to analyze vegetation dynamics over the 1999–2018 based on Landsat normalized combustion ratio time-series. By overlaying the latest vegetation types and spatial distribution of vegetation dynamics, the main types of vegetation change were examined. Results showed that the overall accuracy and kappa coefficient of adaptive-stacking classification were 94.59% and 0.92, respectively, which were higher than those of support vector machine and random forest. The overall accuracy of change detection for three types (vegetation gain, vegetation loss, and no changes) was 83.67%. In the past 20 years, 2,604.43 and 5,458.84 km² land has experienced vegetation loss and gain, respectively. The increase in the areas of forest, reed, and sedge in wetland were 1330.39, 86.42, 136.97 km², respectively. We found that the overall recovery condition of the wetland vegetation around Dongting Lake was good, which demonstrated the key role played by the national wetland ecological protection policy.
... The high-precision classification and rapid monitoring of wetland vegetation is an important basis for the systematic study of the structure and ecological function of wetlands, and provides necessary reference information for the protection and rational development of wetlands. The existence of remote sensing technology not only solves the obstacle of field investigation caused by the inaccessibility of wetland environment, but also makes the monitoring of wetland vegetation more timely and effective (Taddeo et al., 2019;Abeysinghe et al., 2019;Rapinel et al., 2019;Jia et al., 2021;Shen et al., 2021). As more and more sensors were deployed on satellite and airborne remote sensing platforms, remote sensing images of the same geographic location have also increased over time, and at the same time, more and more algorithms for identifying and classifying remote sensing images were used. ...
Article
Full-text available
The verification of the transfer learning ability of the convolutional neural network in the classification of natural vegetation is relatively lacking. In this paper, 16 combination scenarios of multispectral images in marsh vegetation were constructed. The influence of the combination with different spatial resolution gradients and spectral dimensions on the classification accuracy of marsh vegetation was systematically studied. Multi-sensor images were used to evaluate the transfer learning ability of DeepLabV3+ and HRNet algorithms in marsh vegetation, and analyse the transfer learning effect of two algorithms in different spatial resolution and spectral ranges. The majority voting method was used to fuse the classifications of high, medium and low spatial resolution images. Based on the largest area method, the fusion results were integrated with multi-scale segmentation to explore the classification ability of the integration of pixel-based classification and object-based classification. The average accuracies of different spatial resolutions to vegetation in multispectral images were statistically analysed in order to quantitatively study the classification ability of spatial resolution to marsh vegetation. The results indicated that: (1) image combination improved the classification accuracy of marsh vegetation in low-resolution images, and decreased the classification accuracy of vegetation in high- and medium-resolution images based on DeepLabV3+ and HRNet algorithms; (2) when GF-1 and Sentinel-2A images were used for combination, the spectral range increased by 1565–1655 nm and 2100–2280 nm from 450–900 nm, the classification accuracy of GF-1 image improved by 0.93–1.77%, and the classification accuracy of Sentinel-2A image decreased by 2.34–4.15%; (3) DeepLabV3+ and HRNet algorithms both have good transfer learning capabilities in the classification of marsh vegetation, but the transfer learning ability was better in images with different spatial resolutions in similar spectral ranges than in images with different spectral ranges; (4) the integration of object-based segmentation and pixel-based classifications (DeepLabV3+ and HRNet algorithms) improved the accuracy, and the growth rate of overall accuracy reached 4.45–5%; (5) the classification of water in images with different spatial resolution gradients had the largest difference, and the accuracy of deep-water marsh vegetation was lower than that of shrub and shallow-water marsh vegetation.
... The efficient wetland category describes ecosystem properties (Merot et al. 2006). In wetland monitoring and management, this categorization can deliver information on wetland loss and on changes in wetland functional categories and was adopted in a remote sensing based study of wetlands in north-eastern France (Rapinel et al. 2019). Accordingly, the basis of our wetland characterization framework consists of a potential wetlands map (potential wetlands), from which actual wetland boundaries are derived (existing wetlands). ...
Article
Full-text available
Wetlands are abundant across the African continent and provide a range of ecosystem services on different scales but are threatened by overuse and degradation. It is essential that national governments enable and ensure the sustainable use of wetland resources to maintain these services in the long run. As informed management decisions require reliable, up-to-date, and large coverage spatial data, we propose a modular Earth observation-based framework for the geo-localisation and characterization of wetlands in East Africa. In this study, we identify four major challenges in spatial data supported wetland management and present a framework to address them. We then apply the framework comprising Wetland Delineation, Surface Water Occurrence, Land Use/Land Cover classification and Wetland Use Intensity for the whole of Rwanda and evaluate the ability of these layers to meet the identified challenges. The layers' spatial and temporal characteristics make them combinable and the information content, of each layer alone as well as in combination, renders them useful for different wetland management contexts.
... Sentinel-1A is broad in scope, and it is possible to extract information on partially or wholly wet meadows using the Shannon Entropy method. Rapinel et al. (2019) developed an alternative method of wetland mapping based on free remote sensing data. The data used are Sentinel-1, Sentinel-2 image data, validated MODIS data with LIDAR data. ...
Article
Full-text available
Ogan Komering Ilir (OKI) Regency area, dominated by wetlands, requires appropriate land use planning to support welfare and environmental conservation programs. This article explores the utilization of Synthetic Aperture Radar (SAR) Sentinel-1A satellite data to identify wetland areas with various land covers to provide a medium scale wetland map quickly. The soil moisture maps were captured from the SEPAL platform to monitor soil moisture patterns throughout 2015–2017. The backscattering profiles of VV (δVV) polarization and VH polarization (δVH) were then analyzed for data sentinel-1 from 2015 to 2018. The ratio between VV and VH (δVV/δVH) values and textures of VV and VH is used as a reference to determine the wetland area; determined on an experimental analysis of soil moisture profile, dielectric constants, and backscattering coefficient values from other studies. The low δVH/δVV ratio (ranges between 1.3 and 1.6) is located in the northern OKI regency and is mainly covered by acacia plantations. These estimations align with the Peat Hydrology Unit maps and ground check as validation data. Overall, the δVH/δVV ratio is relatively the same yearly, with a small variation value. This study found the capability of Sentinel-1 for the accurate classification of wetland vegetation types, primarily herbaceous or shrubby vegetated wetlands, but not for high-vegetated wetlands.
... To perform periodic wetlands monitoring, remote sensing has become an essential tool, as it is shown in references [9] [10]. Over the years, remote sensing platforms have been successfully tested for this application, from satellites [11] to UAVs [12] [13] .However, every remote sensing technique presents different constraints depending on the operating conditions, in the case of high altitude wetlands (Andean), there are many limitations for their implementation due to the cloudiness, accessibility, and wind gusts of the region. ...
Article
Full-text available
Remote sensing using satellites and unmanned aerial vehicles (UAVs) has become an important tool for wetland delimitation and saturation assessment since they enable patterns identification and wetland saturation data collection in an agile and optimum way. However, their deployment and operative costs limit their implementation in harsh environments, such as the ones presented in the high Andean wetlands. In this context, this work presents a framework to monitor cost-effectively high Andean wetlands using a multi-agent approach based on: field testing, UAV orthomosaics, and satellite imagery. The method developed comprises two stages: i) definition of the monitoring agent (field testing, satellite, UAV) and ii) image processing. For these stages, semi-empirical and statistical models, which were developed in previous works are incorporated in an open-source framework to tailor each monitoring approach accordingly to the seasonality of a representative Andean wetland. The application of the method and its results highlight the suitability of using visual spectrum low-cost remote sensing approach to compute wetlands saturation percentage. In addition, the methodology proposed allowed the development of a temporal monitoring scheme, where the viability of each monitoring agent is examined. In order to validate the method, field data and multispectral imagery were employed using as case of study the Pugllohuma wetland located in the Antisana Reserve. Thus, the main contribution of this work lies in establishing a technified monitoring framework for the Ecuadorian high Andean wetlands, which can be scaled up and extrapolated to other wetlands with similar harsh environmental conditions, helping to their management and protection policies decision-making.
... Beyond these broad-scale maps based on the CORINE Land Cover layer, many studies based on automatic and fine-scale analyses have demonstrated the contribution of multi-temporal and high-spatial-resolution satellite data in discriminating grasslands from other LULC types [16,[23][24][25], characterizing forage quality [20], identifying agricultural practices [26] and mapping floristic variation in semi-natural grasslands [27][28][29]. However, discriminating semi-natural and temporary grasslands accurately remains a concern [23,30] due to the lack of temporal depth in remote sensing time-series and because a one-year observation is insufficient to discriminate between semi-natural and temporary grasslands [31]. For instance, the French national land use map as well as the European high-resolution layer (HRL) for "grassland", both derived from multi-temporal Sentinel and Landsat data, combine semi-natural and temporary grasslands into a single "grassland" class. ...
Article
Full-text available
Semi-natural grasslands are perennial ecosystems and an important part of agricultural landscapes that are threatened by urbanization and agricultural intensification. However, implementing national grassland conservation policies remains challenging because their inventory, based on short-term observation, rarely discriminate semi-natural permanent from temporary grasslands. This study aims to map grassland frequency at a national scale over a long period using Moderate Resolution Imaging Spectroradiometer (MODIS) 250 m satellite time-series. A three-step method was applied to the entire area of metropolitan France (543,940 km²). First, land-use and land-cover maps—including grasslands—were produced for each year from 2006–2017 using the random forest classification of MOD13Q1 and MYD13Q1 products, which were calibrated and validated using field observations. Second, grassland frequency from 2006–2017 was calculated by combining the 12 annual maps. Third, sub-pixel analysis was performed using a reference layer with 20 m spatial resolution to quantify percentages of land-use and land-cover classes within MODIS pixels classified as grassland. Results indicate that grasslands were accurately modeled from 2006–2017 (F1-score 0.89–0.93). Nonetheless, modeling accuracy varied among biogeographical regions, with F1-score values that were very high for Continental (0.94 ± 0.01) and Atlantic (0.90 ± 0.02) regions, high for Alpine regions (0.86 ± 0.04) but moderate for Mediterranean regions (0.62 ± 0.10). The grassland frequency map for 2006–2017 at 250 m spatial resolution provides an unprecedented view of stable grassland patterns in agricultural areas compared to existing national and European GIS layers. Sub-pixel analysis showed that areas modeled as grasslands corresponded to grassland-dominant areas (60%–94%). This unique long-term and national monitoring of grasslands generates new opportunities for semi-natural grassland inventorying and agro-ecological management.
Article
Over the past decade, open source software has become widely accepted across governments, industries and academia. The geospatial domain is no exception and this trend is also reflected in geospatial research and practice. Nowadays, governments and stakeholders from the business sector both participate and promote open geospatial science including open geospatial data and open source geospatial software. As a result, open source geospatial science and software (i.e., open source GIS) is a growing area of research with numerous applications and great potential. The consistent prevalence of open source GIS studies motivated this thematic collection. The contributions are divided into two main categories. In the first, novel open source geospatial software and standards are presented, each of which has been implemented for and applied to a certain use case, and at the same time may be applied to other use cases due to the reproducibility of open source software. The second category presents and discusses the applicability and usability of open source GIS solutions for various interdisciplinary domains, mostly related to urban studies.
Article
Coastal wetlands play a vital role in coastline protection. Coastal landscape of the Guangdong-Hong Kong-Macao Greater Bay Area (GBA) has changed drastically. In order to explore the process and law of coastal landscape dynamics in GBA, in this research, multi-temporal Landsat images and proposed hybrid object-based image analysis and multi-level decision tree classification were applied to monitor the changing trends and current status of the coastal landscape in GBA from 1988 to 2018. There were 5818.58 km² of natural wetlands in 2018, including mangroves (39.62 km²), tidal flats (73.64 km²), rivers (161.65 km²), and shallow seas (5543.67 km²). Due to the influence of anthropogenic activities, natural wetlands showed a decreasing trend, while artificial wetlands and non-wetlands have been increasing. Results indicated that the landscape in the study area showed a trend of fragmentation, the degree of landscape aggregation weakened, the dominant patches were scattered, the shape became more and more complex, and the landscape heterogeneity increased. Then, a future land use simulation model was used to simulate the coastal landscape change trends of GBA in 2028 under four scenarios. The mangrove has the fastest growth rate of 1.47 km²/a under wetland protection scenario. The tidal flat and paddy field would increase under scenarios of the wetland protection and farmland protection, respectively. The construction land has the fastest growth rate under the economic development scenario and would be 47.56 km²/a. Woodland, dry land, and culture ponds are further transformed into other landscape types, while rivers, shallow seas, and reservoirs/ponds have no obvious change.
Article
Transparent, consistent, and statistically reliable land use/ land cover area estimates are needed to assess land use change and greenhouse gas emissions associated with biofuel production and other land uses that are influenced by policy. As relevant studies have increased rapidly during past decades, the methods used to combine data extracted from land use land cover (LULC) surveys and remote sensing-based products and track or report sources of uncertainty vary notably. This paper reviews six data sources that are most commonly used to investigate LULC and change in the contiguous U.S. by highlighting the main characteristics, strengths and weaknesses and considering how uncertainty is assessed by the June Area Survey (JAS), the Census of Agriculture (COA), the Farm Survey Agency (FSA) acreage, the National Resources Inventory (NRI), the National Wetlands Inventory (NWI), and the Forest Inventory and Analysis (FIA); and two remote sensing-based data products, the Cropland Data Layer (CDL) and the National Land Cover Database (NLCD). The summary and conclusion identify important research gaps or challenges limiting current land use/land cover and change studies (e.g., lack of high-quality reference data and uncertainty quantification, etc.) and opportunities and emerging techniques (data fusion and machine learning) that will improve reliability of land use/land cover assessments and associated policies. Blended approaches that marry high quality ground truth data that are more finely resolved than data supplied by government surveys with multitemporal imagery are needed track use of non-agricultural lands vulnerable to agricultural expansion. These considerations are notably important as the U.S. considers the renewal and possibly revision of its Renewable Fuel Standard, which includes provisions that require monitoring of agricultural land expansion.
Article
Present study has attempted to measure Water Richness (WR) and Wetland Habitat Suitability (WHS) in deltaic environment and assessed their spatial linkages. Water richness exhibits availability of water in wetland and its dynamicity, whereas wetland habitat suitability depicts physical habitat ambiance of a wetland toward vibrant ecosystem. Both the components are very essential and should be measured to explore ecosystem service and environmental heath of a region. For investigating water richness of the wetland six water availability indicating parameters have been chosen and for assessing wetland habitat suitability four additional parameters have been taken into consideration. Four widely used and recognised machine learning algorithms like Reduced Error Pruning (REP) tree, Random forest, Artificial Neural Network (ANN) and Support Vector Machine (SVM) have been employed here in order to develop suitable model at two phases. Results reveal that very high water rich zone is found over 200–215 km² wetland area followed by high water rich zone over 125–140 km² wetland area in both the phases. Wetland habitat suitability assessment shows only 100–150 km² of the wetland having very high suitability and 110–120 km² of wetland having high suitability. Field investigation and accuracy assessment support the validity and acceptability of the results. Spatial linkage between water richness and habitat suitability demonstrates that 30–40% very high water rich zone represents very high habitat suitability figuring out importance of both the models. Therefore, results recommend that only water richness of the wetlands of the wetlands is not enough to represent the habitat suitability in the densely populated riparian flood plain region.
Article
We use a Wiener deconvolution filter to deblur the streak image of streak tube imaging lidar to improve the spatial resolution of the system and reduce the edge blurring effect in point clouds. Experiments were performed to investigate the performance of the deconvolution method. Results show that the spatial resolution improved from 9 to 4.5 mm, and the root-mean-square errors of the edge regions are effectively reduced. Additionally, the transition section decreases from 14 to 5.6 mm when the target is 5 m away from the receiver.
Article
Full-text available
Wetlands are one of the most important ecosystems that provide a desirable habitat for a great variety of flora and fauna. Wetland mapping and modeling using Earth Observation (EO) data are essential for natural resource management at both regional and national levels. However, accurate wetland mapping is challenging, especially on a large scale, given their heterogeneous and fragmented landscape, as well as the spectral similarity of differing wetland classes. Currently, precise, consistent, and comprehensive wetland inventories on a national- or provincial-scale are lacking globally, with most studies focused on the generation of local-scale maps from limited remote sensing data. Leveraging the Google Earth Engine (GEE) computational power and the availability of high spatial resolution remote sensing data collected by Copernicus Sentinels, this study introduces the first detailed, provincial-scale wetland inventory map of one of the richest Canadian provinces in terms of wetland extent. In particular, multi-year summer Synthetic Aperture Radar (SAR) Sentinel-1 and optical Sentinel-2 data composites were used to identify the spatial distribution of five wetland and three non-wetland classes on the Island of Newfoundland, covering an approximate area of 106,000 km2. The classification results were evaluated using both pixel-based and object-based random forest (RF) classifications implemented on the GEE platform. The results revealed the superiority of the object-based approach relative to the pixel-based classification for wetland mapping. Although the classification using multi-year optical data was more accurate compared to that of SAR, the inclusion of both types of data significantly improved the classification accuracies of wetland classes. In particular, an overall accuracy of 88.37% and a Kappa coefficient of 0.85 were achieved with the multi-year summer SAR/optical composite using an object-based RF classification, wherein all wetland and non-wetland classes were correctly identified with accuracies beyond 70% and 90%, respectively. The results suggest a paradigm-shift from standard static products and approaches toward generating more dynamic, on-demand, large-scale wetland coverage maps through advanced cloud computing resources that simplify access to and processing of the “Geo Big Data.” In addition, the resulting ever-demanding inventory map of Newfoundland is of great interest to and can be used by many stakeholders, including federal and provincial governments, municipalities, NGOs, and environmental consultants to name a few.
Article
Full-text available
Peatlands that are close to a natural state are rich in biodiversity and are significant carbon storages. Simultaneously, peat resources are of interest to industry, which leads to competing interests and tensions regarding the use and management of peatlands. In this case study, we studied knowledge-management interactions through the development of participation and the resulting representation of nature (how nature was described), as well as the proposed and implemented conservation policy instruments. We focused on the years 2009-2015, when peatland management was intensively debated in Finland. We did an interpretative policy analysis using policy documents (Peatland Strategy; Government Resolution; Proposal for Conservation Programme) and environmental legislation as central data. Our results show how the representation of nature reflected the purpose of the documents and consensus of participants' values. The representation of nature changed from skewed use of ecosystem services to detailed ecological knowledge. However, simultaneously, political power changed and the planned supplementation programme for peatland conservation was not implemented. The Environment Protection Act was reformulated so that it prohibited the use of the most valuable peatlands. Landowners did not have the chance to fully participate in the policy process. Overall, the conservation policy instruments changed to emphasize voluntariness but without an adequate budget to ensure sufficient conservation.
Article
Full-text available
Wetlands play a key role in controlling flooding and non-point-source (diffuse) pollution. They are therefore an important tool for mitigating diffuse water pollution from farms. However, to use this tool, it is necessary to obtain detailed assessments and identification of potential wetland restoration or creation sites. This is complicated by the diversity of landscapes, environmental conditions, and land ownership. Site suitability for wetland restoration or creation depends on many factors: the underlying geology, soils, topography, hydrology, drainage, and land ownership. Local hydrology and soils are among the most important factors. However, the inventory and characterization of a site’s soils and hydrology often requires extensive, expensive, and time-consuming ground surveys, and it is therefore limited to small areas. Another possibility would be to consider topography, which strongly determines water movement patterns. Light detection and ranging (LiDAR) data provides detailed topographic information and can be acquired by remote sensing. Our study showed that terrain analysis using high-resolution topographical data can produce suitability maps for wetlands that can be easily used by decision makers and planners in watershed management. The rapid methodology reveals potential wetland creation or restoration sites at a reasonable cost; with the resulting spatially explicit suitability map, managers can plan for wetland creation or restoration without having to wait for field-data collection.
Chapter
Full-text available
Over the past decades, remote sensing has been repeatedly identified as a promising and powerful tool to aid nature conservation. Many methods and applications of remote sensing to monitor biodiversity have indeed been published, and continue to be at an increasing rate. As such, remote sensing is seemingly living up to its expectations; yet, its actual use in nature conservation (or rather the lack thereof) contradicts this. We argue that, at least for the practical implementation of regular vegetation monitoring, including within protected areas (e.g., European Natura 2000 sites), a lack of transferability of remote sensing methods is an overlooked factor that hinders its effective operational use for nature conservation. Among the causes of poor method transferability is the large variation in objects of interest, user requirements, ground reference data, and image properties, but also the lack of consideration of transferability during method development. To stimulate the adoption of remote sensing based techniques in vegetation monitoring and conservation, we recommend that a number of actions are taken. We call upon remote sensing scientists and nature monitoring experts to specifically consider and demonstrate method transferability by using widely available image data, limiting ground reference data dependence, and making their preferably open-source programming code publicly available. Furthermore, we recommend that nature conservation specialists are open and realistic about potential outcomes by not expecting the replacement of current in-place methodologies, and actively contributing to the thought process of generating transferable and repeatable methods.
Article
Full-text available
Societal, economic and scientific interests in knowing where biodiversity is, how it is faring and what can be done to efficiently mitigate further biodiversity loss and the associated loss of ecosystem services are at an all-time high. So far, however, biodiversity monitoring has primarily focused on structural and compositional features of ecosystems despite growing evidence that ecosystem functions are key to elucidating the mechanisms through which biological diversity generates services to humanity. This monitoring gap can be traced to the current lack of consensus on what exactly ecosystem functions are and how to track them at scales beyond the site level. This contribution aims to advance the development of a global biodiversity monitoring strategy by proposing the adoption of a set of definitions and a typology for ecosystem functions, and reviewing current opportunities and potential limitations for satellite remote sensing technology to support the monitoring of ecosystem functions worldwide. By clearly defining ecosystem processes, functions and services and their interrelationships, we provide a framework to improve communication between ecologists, land and marine managers, remote sensing specialists and policy makers, thereby addressing a major barrier in the field.
Article
Full-text available
Orfeo ToolBox is an open-source project for state-of-the-art remote sensing, including a fast image viewer, applications callable from command-line, Python or QGIS, and a powerful C++ API. This article is an introduction to the Orfeo ToolBox’s flagship features from the point of view of the two communities it brings together: remote sensing and software engineering.
Article
Full-text available
Temporal streams are vitally important for hydrology and riverine ecosystems. The identification of wet channel networks and spatial and temporal dynamics is essential for effective management, conservation, and restoration of water resources. This study investigated the temporal dynamics of stream networks in five watersheds under different climate conditions and levels of human interferences, using a systematic method recently developed for extracting wet channel networks based on light detection and ranging elevation and intensity data. In this paper, thresholds of canopy height for masking densely vegetated areas and the ‘time of forward diffusion’ parameter for filtering digital elevation model are found to be greatly influential and differing among sites. The inflection point of the exceedance probability distribution of elevation differences in each watershed is suggested to be used as the canopy height threshold. A lower value for the ‘time of forward diffusion’ is suggested for watersheds with artificial channels. The properties of decomposed and composite probability distribution functions of intensity and the extracted intensity thresholds are found to vary significantly among regions. Finally, the wet channel density and its variation with climate for five watersheds are found to be reasonable and reliable according to results reported previously in other regions.
Article
Full-text available
Plant primary production is a key driver of several ecosystem functions in seasonal marshes, such as water purification and secondary production by wildlife and domestic animals. Knowledge of the spatio-temporal dynamics of biomass production is therefore essential for the management of resources—particularly in seasonal wetlands with variable flooding regimes. We propose a method to estimate standing aboveground plant biomass using NDVI Land Surface Phenology (LSP) derived from MODIS, which we calibrate and validate in the Doñana National Park’s marsh vegetation. Out of the different estimators tested, the Land Surface Phenology maximum NDVI (LSP-Maximum-NDVI) correlated best with ground-truth data of biomass production at five locations from 2001–2015 used to calibrate the models (R ² = 0.65). Estimators based on a single MODIS NDVI image performed worse (R ² ≤ 0.41). The LSP-Maximum-NDVI estimator was robust to environmental variation in precipitation and hydroperiod, and to spatial variation in the productivity and composition of the plant community. The determination of plant biomass using remote-sensing techniques, adequately supported by ground-truth data, may represent a key tool for the long-term monitoring and management of seasonal marsh ecosystems. View Full-Text
Article
Full-text available
A detailed and accurate knowledge of land cover is crucial for many scientific and operational applications, and as such, it has been identified as an Essential Climate Variable. This accurate knowledge needs frequent updates. This paper presents a methodology for the fully automatic production of land cover maps at country scale using high resolution optical image time series which is based on supervised classification and uses existing databases as reference data for training and validation. The originality of the approach resides in the use of all available image data, a simple pre-processing step leading to a homogeneous set of acquisition dates over the whole area and the use of a supervised classifier which is robust to errors in the reference data. The produced maps have a kappa coefficient of 0.86 with 17 land cover classes. The processing is efficient, allowing a fast delivery of the maps after the acquisition of the image data, does not need expensive field surveys for model calibration and validation, nor human operators for decision making, and uses open and freely available imagery. The land cover maps are provided with a confidence map which gives information at the pixel level about the expected quality of the result.
Article
Full-text available
Wetlands are highly productive and support a wide variety of ecosystem goods and services. Various forms of global change impose compelling needs for timely and reliable information on the status of wetlands worldwide, but several characteristics of wetlands make them challenging to monitor remotely: they lack a single, unifying land-cover feature; they tend to be highly dynamic and their energy signatures are constantly changing; and steep environmental gradients in and around wetlands produce narrow ecotones that often are below the resolving capacity of remote sensors. These challenges and needs set the context for a special issue focused on wetland remote sensing. Contributed papers responded to one of three overarching questions aimed at improving remote, large-area monitoring of wetlands: (1) What approaches and data products are being developed specifically to support regional to global long-term monitoring of wetland landscapes? (2) What are the promising new technologies and sensor/multisensor approaches for more accurate and consistent detection of wetlands? (3) Are there studies that demonstrate how remote long-term monitoring of wetland landscapes can reveal changes that correspond with changes in land cover and land use and/or changes in climate?
Article
Full-text available
Comprehensive wetland inventories are an essential tool for wetland management, but developing and maintaining an inventory is expensive and technically challenging. Funding for these efforts has also been problematic. Here we describe a large-area application of a semi-automated process used to update a wetland inventory for east-central Minnesota. The original inventory for this area was the product of a labor-intensive, manual photo-interpretation process. The present application incorporated high resolution, multi-spectral imagery from multiple seasons; high resolution elevation data derived from lidar; satellite radar imagery; and other GIS data. Map production combined image segmentation and random forest classification along with aerial photo-interpretation. More than 1000 validation data points were acquired using both independent photo-interpretation as well as field reconnaissance. Overall accuracy for wetland identification was 90 % compared to field data and 93 % compared to photo-interpretation data. Overall accuracy for wetland type was 72 and 78 % compared to field and photo-interpretation data, respectively. By automating the most time consuming part of the image interpretations, initial delineation of boundaries and identification of broad wetland classes, we were able to allow the image interpreters to focus their efforts on the more difficult components, such as the assignment of detailed wetland classes and modifiers.
Article
Full-text available
The correction of atmospheric effects is one of the preliminary steps required to make quantitative use of time series of high resolution images from optical remote sensing satellites. An accurate atmospheric correction requires good knowledge of the aerosol optical thickness (AOT) and of the aerosol type. As a first step, this study compares the performances of two kinds of AOT estimation methods applied to FormoSat-2 and LandSat time series of images: a multi-spectral method that assumes a constant relationship between surface reflectance measurements and a multi-temporal method that assumes that the surface reflectances are stable with time. In a second step, these methods are combined to obtain more accurate and robust estimates. The estimated AOTs are compared to in situ measurements on several sites of the AERONET (Aerosol Robotic Network). The methods, based on either spectral or temporal criteria, provide accuracies better than 0.07 in most cases, but show degraded accuracies in some special cases, such as the absence of vegetation for the spectral method or a very quick variation of landscape for the temporal method. The combination of both methods in a new spectro-temporal method increases the robustness of the results in all cases.
Article
Full-text available
Satellite remote sensing is an important tool for monitoring the status of biodiversity and associated environmental parameters, including certain elements of habitats. However, satellite data are currently underused within the biodiversity research and conservation communities. Three factors have significant impact on the utility of remote sensing data for tracking and understanding biodiversity change. They are its continuity, affordability, and access. Data continuity relates to the maintenance of long-term satellite data products. Such products promote knowledge of how biodiversity has changed over time and why. Data affordability arises from the cost of the imagery. New data policies promoting free and open access to government satellite imagery are expanding the use of certain imagery but the number of free and open data sets remains too limited. Data access addresses the ability of conservation biologists and biodiversity researchers to discover, retrieve, manipulate, and extract value from satellite imagery as well as link it with other types of information. Tools are rapidly improving access. Still, more cross-community interactions are necessary to strengthen ties between the biodiversity and remote sensing communities. (C) 2014 Published by Elsevier Ltd.
Article
Full-text available
Over the last 30 years, ecological networks have been deployed to reduce global biodiversity loss by enhancing landscape connectivity. Bird species dwelling in woodland habitats that are embedded in agriculture-dominated landscapes are expected to be particularly sensitive to the loss of connectivity. This study aimed to determine the role of landscape connectivity in woodland bird species richness, abundance, and community similarity in north-east Brittany (north-west France). An exhaustive woodland selection protocol was carried out to minimize the effects of woodland size on the response variables. Connectivity of the woodland and forest network in the study area was evaluated using graph-theory, accounting for matrix permeability, and a characteristic median natal dispersal distance at the community level based on the bird species pool recorded in the sampled woodlands. Information-theoretic model selection, controlling for woodland size in all the cases, depicted the response of woodland birds at the community level to the connectivity of agriculture-dominated landscapes. On average, the sampled woodlands (n = 25) contained 15.5 ± 2.4 bird species, with an abundance of 25.1 ± 3.9, and had highly similar bird communities (species composition and proportion); eight species represented 57% of total abundance and were present in at least 22 woodlands. The performance of models improved when using effective, rather than Euclidean, interpatch distances in the connectivity assessment. Landscape connectivity was only significantly related to similarity of proportional species composition. Large woodlands contained communities with more similar species proportions in an inhospitable agricultural landscape matrix than in a more permeable one. Woodland size was the most relevant factor determining species abundance, indicating that the bird population sizes are primarily proportional to the local habitat availability. Connectivity in relation to landscape matrix permeability did not seem to induce the flow of woodland-dependent bird species that are dominant in the community but rather of matrix-dwelling bird species that are less dependent on woodland patch area. In conclusion, both habitat conservation and restoration (i.e., amount and quality), in combination with permeable landscape structures (such as heterogeneous land cover mosaics), are advocated for community level conservation strategies.
Book
Full-text available
Wetlands perform functions that deliver benefits to society, often referred to as ecosystem services. These ecosystem services include water supply, flood regulation, water purification, climate regulation, biodiversity, agriculture (e.g. grazing land), and amenity. A functional approach to wetland assessment enables a holistic view to be taken of the wide range of services wetlands can provide. The functional assessment procedures (FAPs) in this volume translate best available scientific knowledge into reasonable predictions of how component parts of wetlands function in different landscape contexts. They can be used to indicate the potential and priorities for management options in such areas as flood control, pollution reduction and biodiversity conservation. Functional assessment enables the user to predict the functioning of a wetland area without the need for comprehensive and expensive empirical research The FAPs therefore provide a methodology that can be used by both experts and non-experts to assess wetland functioning relatively rapidly. The volume includes an electronic version of the FAPs on CD which automates aspects of the assessment once the initial recording stage is completed. It is anticipated that the FAPs will be used by a range of individuals or organisations concerned with wetland management who wish to gain a better understanding of the processes, functions, services or benefits and potential of the wetlands for which they have responsibility. Provides a systematic methodology to evaluate how wetlands function. Allows on-experts to assess wetland functioning rapidly and cost-effectively. Automates aspects of the functional assessment through the accompanying CD-ROM.
Article
Full-text available
With increasing scarcity of natural resources, there is a need to provide resource managers and planners with maps that reliably inform about areas vulnerable to hydrological risks, including areas with ephemeral to intermittent flows. This paper demonstrates that the newly developed Wet-Areas Mapping (WAM) process using LiDAR-based point cloud data addresses some of these needs. This is done by portraying local flow patterns, soil drainage, soil moisture regimes and natural vegetation type across mapped areas in a numerically robust and consistent manner. As a result, WAM-derived maps are useful for “surprise-free” operations planning in several areas of natural resource planning (forestry, parks and recreation, oil and gas extraction, land reclamation), and also serve as field guides for locating and delineating flow channels, road-stream crossings, wet areas and wetlands.
Article
Full-text available
Low saltmarshes are subjected to variability between sites and tidal cycles in terms of erosive forces by current and waves, the frequency and duration of flooding and soil salinity. The establishment of vegetation in pioneer zones is directly related to sedimentary dynamics but few data are available concerning the effects of plants on sediment dynamics. In the Mont Saint Michel Bay (France), the low saltmarshes, including pioneer zones, are characterized by a micro-topography composed of hummocks with vegetation dominated by Puccinellia maritima, mudflats with a low sparse vegetation of Spartina anglica, Salicornia fragilis and Puccinellia maritima and a few erosion zones. The aim of this study was to (1) investigate the sediment deposition and soil elevation patterns, between tidal cycles and between sites; (2) look for a relationship between the development and dynamics of the micro-topography and the different plant species; and (3) evaluate whether Puccinellia maritima plays any role in enhancing sediment deposition and therefore plant succession in these lower marshes.
Article
Full-text available
Ecosystem services are natural assets produced by the environment and utilized by humans -- such as clean air, water, food and materials -- and contribute to social and cultural well-being. This concept, arguably, has been developed further in wetlands than any other ecosystem. Wetlands were historically important in producing the extensive coal deposits of the Carboniferous period; key steps in human development took place in communities occupying the wetland margins of rivers, lakes and the sea; and wetlands play a key role in the hydrological cycle influencing floods and river droughts. In this paper we examine three pillars that support the wetland research agenda: hydrology, wetland origins and development, and linkages to society. We investigate these through an overview of the evolution of wetland science and assessment of the wide range of topics relating to ecosystem services covered in this Special Issue. We explain the seminal change in how modern society values the benefits of natural ecosystems and highlight the pathfinder role that wetland research has played in the paradigm shift.Co-editors D. Koutsoyiannis and Z.W. KundzewiczCitation Maltby, E. and Acreman, M.C., 2011. Ecosystem services of wetlands: pathfinder for a new paradigm. Hydrological Sciences Journal, 56 (8), 1341–1359.
Article
Full-text available
An important goal of conservation biology is the maintenance of ecosystem processes. Incorporating quantitative measurements of ecosystem functions into conser-vation practice is important given that it provides not only proxies for biodiversity patterns, but also new tools and criteria for management. In the satellite era, the translation of spectral information into ecosystem functional variables expands and complements the more traditional use of satellite imagery in conservation biology. Remote sensing scientists
Article
Full-text available
Aim To examine the geographical patterns of the interception of photosynthetically active radiation by vegetation and to describe its spatial heterogeneity through the definition of ecosystem functional types (EFTs) based on the annual dynamics of the Normalized Difference Vegetation Index (NDVI), a spectral index related to carbon gains. Location The Iberian Peninsula. Methods EFTs were derived from three attributes of the NDVI obtained from NOAA/AVHRR sensors: the annual integral (NDVI-I), as a surrogate of primary production, an integrative indicator of ecosystem functioning; and the intra-annual relative range (RREL) and month of maximum NDVI (MMAX), which represent key features of seasonality. Results NDVI-I decreased south-eastwards. The highest values were observed in the Eurosiberian Region and in the highest Mediterranean ranges. Low values occurred in inner plains, river basins and in the southeast. The Eurosiberian Region and Mediterranean mountains presented the lowest RREL, while Eurosiberian peaks, river basins, inner-agricultural plains, wetlands and the southeastern part of Iberia presented the highest. Eurosiberian ecosystems showed a summer maximum of NDVI, as did high mountains, wetlands and irrigated areas in the Mediterranean Region. Mediterranean mountains had autumn–early-winter maxima, while semi-arid zones, river basins and continental plains had spring maxima. Based on the behaviour in the functional traits, 49 EFTs were defined. Main conclusions The classification, based on only the NDVI dynamics, represents the spatial heterogeneity in ecosystem functioning by means of the interception of radiation by vegetation in the Iberian Peninsula. The patterns of the NDVI attributes may be used as a reference in evaluating the impacts of environmental changes. Iberia had a high spatial variability: except for biophysically impossible combinations (high NDVI-I and high seasonality), almost any pattern of seasonal dynamics of radiation interception was represented in the Peninsula. The approach used to define EFTs opens the possibility of monitoring and comparing ecosystem functioning through time.
Article
Full-text available
Question: What is the influence of management on the functioning of vegetation over time in Mediterranean ecosystems under different climate conditions? Location: Mediterranean shrublands and forests in SE Iberia (Andalusia). Methods: We evaluated the Normalized Difference Vegetation Index (NDVI) for the 1997-2002 time series to determine phenological vegetation patterns under different historical management regimes. Three altitudinal ranges were considered within each area to explore climate × management interactions. Each phenological pattern was analysed using time series statistics, together with precipitation (monthly and cumulative) and temperature. Results: NDVI time series were significantly different under different management regimes, particularly in highly transformed areas, which showed the lowest NDVI, weakest annual seasonality and a more immediate phenological response to precipitation. The NDVI relationship with precipitation was strongest in the summer-autumn period, when precipitation is the main plant growth-limiting factor. Conclusions: NDVI time series analyses elucidated complex influences of land use and climate on ecosystem functioning in these Mediterranean ecosystems. We demonstrated that NDVI time series analyses are a useful tool for monitoring programmes because of their sensitivity to changes, ease of use and applicability to large-scale studies.
Article
Full-text available
Local species coexistence is the outcome of abiotic and biotic filtering processes which sort species according to their trait values. However, the capacity of trait-based approaches to predict the variation in realized species richness remains to be investigated. In this study, we asked whether a limited number of plant functional traits, related to the leaf-height-seed strategy scheme and averaged at the community level, is able to predict the variation in species richness over a flooding disturbance gradient. We further investigated how these mean community traits are able to quantify the strength of abiotic and biotic processes involved in the disturbance–productivity–diversity relationship. We thus tested the proposal that the deviation between the fundamental species richness, assessed from ecological niche-based models, and realized species richness, i.e. field-observed richness, is controlled by species interactions. Flooding regime was determined using a detailed hydrological model. A precise vegetation sampling was performed across 222 quadrats located throughout the flooding gradient. Three core functional traits were considered: specific leaf area (SLA), plant height and seed mass. Species richness showed a hump-shaped response to disturbance and productivity, but was better predicted by only two mean community traits: SLA and height. On the one hand, community SLA that increased with flooding, controlled the disturbance-diversity relationship through habitat filtering. On the other hand, species interactions, the strength of which was captured by community height values, played a strong consistent role throughout the disturbance gradient by reducing the local species richness. Our study highlights that a limited number of simple, quantitative, easily measurable functional traits can capture the variation in plant species richness at a local scale and provides a promising quantification of key community assembly mechanisms.
Article
En conditions climatiques tempérées, dans des contextes géomorphologiques avec substrat à faible profondeur et faible perméabilité, à pentes modérées, la nappe est généralement proche de la surface du sol en bas de versant.. Ces conditions conduisent de façon saisonnière à la présence de petites zones humides ripariennes de quelques hectares au plus. Ces zones sont insérées et dispersées au sein de paysages agricoles. Elles sont souvent oubliées des inventaires des zones humides bien qu'elles jouent un rôle important dans le contrôle de l'hydrologie et de la qualité des eaux des bassins versants. Une typologie hydrologique de ces petites zones humides est proposée ici pour accompagner la réflexion sur leur gestion raisonnée, confrontée à des objectifs antagonistes de maintien de biodiversité et de lutte contre la pollution.
Article
As one of the most important ecosystems, wetlands are threatened from both natural and anthropogenic activities. Mapping wetland is one of the curtail needs in order to prevent further loss. Since the beginning of the Remote Sensing revolution, different approaches using satellite images have been used for mapping and monitoring wetlands. In this paper we investigate the potential of the recently launched Sentinel satellites, both separate and in combination, for accurately mapping of different wetland classes using Support Vector Machines (SVMs) learning classifier. For investigating the influence of the Sentinel-2 red-edge bands, and the radar bands from Sentinel-1, three different datasets have been analyzed. The results showed that for more accurate mapping of different wetland classes, different datasets should be used. Thus, the red-edge bands have significant influence over the intensive vegetated wetland classes such as swamps, and the radar bands have significant influence over partially decayed vegetated wetland areas such as bogs. For future studies, in addition to the analyzed datasets, we recommend adding and investigating several vegetation indices for mapping and monitoring wetland areas.
Article
Dry grasslands are species rich and ecologically valuable habitats that have experienced a massive decline in Switzerland during the last century due to agricultural intensification and land abandonment. Appropriate management is a key factor in maintaining habitat quality of the remaining most valuable sites and should thus be an essential part of monitoring studies. However, information on management is often missing and fine-scale patterns are difficult to assess, especially over large areas and for past decades. The aim of this study was to predict habitat quality of protected dry grasslands in Switzerland. Using a nation-wide in-situ vegetation data set with plot-based species lists, we derived six habitat quality indicators (management tolerance, light availability, nutrient content, moisture content and species richness). We then tested how well satellite-based phenology metrics, in combination with environmental and climate data, can predict these dry grassland habitat quality indicators. We expected that the seasonal pattern of vegetation activity, based on the Normalized Difference Vegetation Index (NDVI), would represent local productivity and management patterns, two crucial indicators of dry grassland habitat quality. Linear regression analysis was conducted to assess the relative importance and ecological relationship of different NDVI metrics and other environmental and climate predictors for habitat quality. Variance partitioning was applied to assess model contributions of the three variable groups which represent different data sources for productivity and management. Accuracies for the habitat quality prediction models ranged between 34% and 57% and significant correlations with multiple NDVI metrics were found. Including NDVI phenology improved all models by 7–12%. Single contributions of NDVI phenology were highest for management tolerance and nutrient content. However, we found high variation of contributions between management types. NDVI metrics were highly informative for the habitat qualities of abandoned sites, but grazing and mowing reduced or even cancelled their predictive power. Moreover, our results demonstrate the limitation of single-date NDVI values in predicting habitat quality of dry grasslands, in particular pastures and meadows. For monitoring applications of dry grasslands, we propose using a combination of NDVI metrics, as our results showed that they greatly improve prediction results of essential habitat qualities. The Landsat legacy dataset facilitates the assessment of habitat changes during past decades and can be complemented in the future with higher resolution data, such as Sentinel-2, to increase the temporal and spatial resolution so analyses are more appropriate for the typically limited size of dry grassland habitat sites in Switzerland.
Article
Mapping semi-natural grassland has become increasingly important with regard to climate variability, invasive species, and the intensification of land use. At the same time, adequate field data collection is of pivotal importance for national and international reporting obligations, such as the European Habitats Directive. We present a remote-sensing-based monitoring framework for a Natura 2000 site with a heterogeneous composition of different grassland communities, using the Random Forest algorithm. Automated training data selection was successfully implemented based on the Random Forest proximity measure (Overall Accuracy ranging from 77.5–86.5%). RapidEye acquisitions originating from the onset of vegetation (prespring and first spring) and senescence (late summer and first autumn) were identified as important phenological phases for mapping semi-natural grassland communities. The derived probability maps of occurrences for each grassland class captured transitions between grassland communities and are therefore a better approximation of real-world conditions compared to classical, discrete maps.
Article
Machine learning offers the potential for effective and efficient classification of remotely sensed imagery. The strengths of machine learning include the capacity to handle data of high dimensionality and to map classes with very complex characteristics. Nevertheless, implementing a machine-learning classification is not straightforward, and the literature provides conflicting advice regarding many key issues. This article therefore provides an overview of machine learning from an applied perspective. We focus on the relatively mature methods of support vector machines, single decision trees (DTs), Random Forests, boosted DTs, artificial neural networks, and k-nearest neighbours (k-NN). Issues considered include the choice of algorithm, training data requirements, user-defined parameter selection and optimization, feature space impacts and reduction, and computational costs. We illustrate these issues through applying machine-learning classification to two publically available remotely sensed data sets.
Article
Geosynphytosociology deals with the study of combinations of vegetation series – or geosigmeta – within landscape. Its main advantage is to assess conservation status based on vegetation dynamics. However, this field-based approach has not been widely applied, because local surveys are not representative of spatio-temporal landscape complexity, which leads to uncertainties and errors for geosigmeta structural and functional mapping. In this context, satellite time series appear as relevant data for monitoring vegetation dynamics. This article aims to assess the contribution of an annual satellite time series for geosigmeta structural and functional mapping. The study area, which focuses on the French Atlantic coast (4630 km²), includes salt, brackish, sub-brackish and fresh marshes. A structural vegetation map was derived from the classification of an annual time series of 38 MODIS images validated with field surveys. The functional vegetation map was derived from the annual Integral of Normalized Difference Vegetation Index (NDVI-I), as an indicator of above-ground net primary production. Results show that geosigmeta were successfully mapped at a scale of 1:250,000 with an overall accuracy of 82.9%. The geosigmeta functional map highlights a strong gradient from the lowest NDVI-I values in salt marshes to the highest values in fresh marshes.
Article
Wetlands are valuable natural resources that provide many benefits to the environment. Therefore, mapping wetlands is crucially important. Several review papers on remote sensing (RS) of wetlands have been published thus far. However, there is no recent review paper that contains an inclusive description of the importance of wetlands, the urgent need for wetland classification, along with a thorough explanation of the existing methods for wetland mapping using RS methods. This paper attempts to provide readers with an exhaustive review regarding different aspects of wetland studies. First, the readers are acquainted with the characteristics, importance, and challenges of wetlands. Then, various RS approaches for wetland classification are discussed, along with their advantages and disadvantages. These approaches include wetland classification using aerial, multispectral, synthetic aperture radar (SAR), and several other data sets. Different pixel-based and objectbased algorithms for wetland classification are also explored in this study. The most important conclusions drawn from the literature are that the red edge and near-infrared bands are the best optical bands for wetland delineation. In terms of SAR imagery, large incidence angles, short wavelengths, and horizontal transmission and vertical reception polarization are best for detecting of herbaceous wetlands, while small incidence angles, long wavelengths, and horizontal transmission and reception polarization are appropriate for mapping forested wetlands.
Article
Wetland area has decreased in most parts of the world and remains threatened by human pressures. However, wetland loss is difficult to accurately detect, delineate and quantify. While wetland distribution is influenced mainly by landform, LiDAR data provide accurate digital elevation models that can be used to delineate wetlands. Our objective was to map wetland loss at a fine-scale using LiDAR data and historical aerial photographs based on a functional typology that identifies potential, existing and efficient wetlands. The study focused on a 132 km2 site with valley bottom wetlands located in western France. Boundaries of potential wetlands were extracted from a LiDAR-derived Digital Terrain Model that was standardized according to channel network elevation. We identified existing wetlands using interpretation of aerial photographs acquired in 1952, 1978 and 2012. We used multiple correspondence analysis to identify different types of wetland loss. Results show that potential wetlands were successfully delineated at 1:5000 (88–90% overall accuracy) and that 14% of existing wetland area was lost. This highlights the importance of identifying “negotiation areas” where wetland restoration is a priority. The results also reveal two main types of wetland loss based on area, geomorphic context, land cover and period of loss.
Article
Remote sensing is one of the most important tools in ecology and conservation for an effective monitoring of ecosystems in space and time. Hence, a proper training is crucial for developing effective conservation practices based on remote sensing data. In this paper we aim to highlight the potential of open-access data and open-source software and the importance of the inter-linkages between these and remote sensing training, with an interdisciplinary perspective. We will first deal with the importance of open access data by further providing several examples of Free and Open Source Software (FOSS) for a deeper and more critical understanding of remote sensing applications.
Article
National Wetland Inventory (NWI) maps are the most comprehensive wetland maps in the U.S., but NWI maps are now outdated in many regions. A consortium led by Ducks Unlimited is updating the NWI for the Great Lakes/Atlantic Region. Updates are complete for several states but have not been verified extensively with field data. We used wetland maps from 129 on-site wetland delineation projects in three Illinois regions to assess accuracy of original and updated NWI maps. We used ancillary spatial data to characterize areas that were incorrectly classified and identify potential sources of error. Across the three regions, the original NWI omitted 49 % of total wetland area for wetlands greater than 0.2 ha, and 57 % of the area mapped by the NWI was non-wetland. The updated NWI omitted less wetland area (40 % omitted for wetlands larger than 0.2 ha), but only slightly improved errors of commission (55 % of mapped area was non-wetland). Polygons mapped as forested wetlands were less likely to be truly wetlands. Small (<0.06 ha) wetlands were often omitted. Errors reflect ambiguity in the definition of wetlands and technical limitations of the NWI methodology. Due to the high error rates, we recommend further refinement of regional wetland inventories.
Article
The objective of this study is to test a cost-effective, physically based Light Detection and Ranging (LiDAR) classification methodology for wetland and upland land cover types within an area exceeding 1,000 km² in the Boreal Plains, Alberta, Canada. Decision criteria are based on physical attributes of the landscape that influence maintenance of land cover types. Results are compared with 38 geolocated measurement plots at land cover boundaries and transition zones, manual delineation of 2,337 wetlands using photogrammetric methods and publicly available land cover classifications. Results suggest that 57% of LiDAR-based wetland classes correspond with delineated wetlands, whereas 37% occur as errors of commission due to excluded wetlands in the manual delineation and confusion with harvested areas. Comparison of classified edges with plot shows that all classifications underestimate wetland area. Residual differences of the LiDAR-based classification are −0.3 m, on average (compared with measured), and have reduced range of error compared with other methods. Multispectral classifications misclassify up to 2/3 of wetland boundaries as a result of lower-resolution mixed pixels. Therefore, high-resolution maps of terrain morphology and vegetation structure provide an accurate, cost-effective means for characterizing wetland vs. upland forest in areas where LiDAR data are available. 2016
Article
This study investigated the effectiveness of using high resolution data to map wetlands in three ecoregions in Minnesota. High resolution data included multispectral leaf-off aerial imagery and lidar elevation data. These data were integrated using an Object-Based Image Analysis (OBIA) approach. Results for each study area were compared against field and image interpreted reference data using error matrices, accuracy estimates, and the kappa statistic. Producer’s and user’s accuracies were in the range of 92 to 96 percent and 91 to 96 percent, respectively, and overall accuracies ranged from 96-98 percent for wetlands larger than 0.20 ha (0.5 acres). The results of this study may allow for increased accuracy of mapping wetlands efforts over traditional remote sensing methods.
Article
A random forest (RF) classifier is an ensemble classifier that produces multiple decision trees, using a randomly selected subset of training samples and variables. This classifier has become popular within the remote sensing community due to the accuracy of its classifications. The overall objective of this work was to review the utilization of RF classifier in remote sensing. This review has revealed that RF classifier can successfully handle high data dimensionality and multicolinearity, being both fast and insensitive to overfitting. It is, however, sensitive to the sampling design. The variable importance (VI) measurement provided by the RF classifier has been extensively exploited in different scenarios, for example to reduce the number of dimensions of hyperspectral data, to identify the most relevant multisource remote sensing and geographic data, and to select the most suitable season to classify particular target classes. Further investigations are required into less commonly exploited uses of this classifier, such as for sample proximity analysis to detect and remove outliers in the training samples.
Article
The Prairie Pothole Region of North America is characterized by numerous, small, wetland depressions that perform important ecological and hydrological functions. Recent studies have shown that total wetland area in the region is decreasing due to cumulative impacts related to natural and anthropogenic changes. The impact of wetland losses on landscape hydrology is an active area of research and management. Various spatially distributed hydrologic models have been developed to simulate effects of wetland depression storage on peak river flows, frequently using dated geospatial wetland inventories. We describe an innovative method for identifying wetland depressions and quantifying their nested hierarchical bathymetric/topographic structure using high-resolution light detection and ranging (LiDAR) data. This contour tree method allows identified wetland depressions to be quantified based on their dynamic filling-spilling-merging hydrological processes. In addition, wetland depression properties, such as surface area, maximum depth, mean depth, storage volume, etc., can be computed for each component of a depression as well as the compound depression. We successfully applied the proposed method to map wetland depressions in the Little Pipestem Creek watershed in North Dakota. The methods described in this study will provide more realistic and higher resolution data layers for hydrologic modeling and other studies requiring characterization of simple and complex wetland depressions, and help prioritize conservation planning efforts for wetland resources.
Article
The project VegFrance is presently designed in order to elaborate a national vegetation database covering all vegetation types and regions. Launched in January 2012, this project is commonly led by Research Institutions as National Center for Scientific Research (UMS3468, BBEES & UMR 6553 ECOBIO), the Museum National Histoire Naturelle (Service du Patrimoine naturel), Fédération Conservatoires Botaniques Nationaux, the French association for Phytosociology and the Ministry of Ecology. The projected database will integrate three main types of dataset: syntaxa reflecting the national classification, relevés describing the vegetation at the landscape level (i.e. synphytosociological relevés) and analytical relevés or any plots which properly describe vegetation. This project is developed in strong connection with the European Vegetation Survey and with the national project for producing a vegetation map (CarHab).
Chapter
IntroductionUnderstanding Wetlands: A Moving TargetAccounting for Spatial and Temporal Variability in WetlandsCompensating for Influences from the Surrounding LandscapeRestoring and Creating WetlandsDepending on WetlandsConclusion AcknowledgementsReferences
Article
Mapping wetlands across both natural and human-altered landscapes is important for the management of these ecosystems. Though they are considered important landscape elements providing both ecological and socioeconomic benefits, accurate wetland inventories do not exist in many areas. In this study, a multi-scale geographic object-based image analysis (GEOBIA) approach was employed to segment three high spatial resolution images acquired over landscapes of varying heterogeneity due to humandisturbance to determine the robustness of this method to changing scene variability. Multispectral layers, a digital elevation layer, normalized-difference vegetation index (NDVI) layer, and a first-order texture layer were used to segment images across three segmentation scales with a focus on accurate delineation of wetland boundaries and wetland components. Each ancillary input layer contributed to improving segmentation at different scales. Wetlands were classified using a nearest neighbor approach across a relatively undisturbed park site and an agricultural site using GeoEye1 imagery, and an urban site using WorldView2 data. Successful wetland classification was achieved across all study sites with an accuracy above 80%, though results suggest that overall a higher degree of landscape heterogeneity may negatively affect both segmentation and classification. The agricultural site suffered from the greatest amount of over and under segmentation, and lowest map accuracy (kappa: 0.78) which was partially attributed to confusion among a greater proportion of mixed vegetated classes from both wetlands and uplands. Accuracy of individual wetland classes based on the Canadian Wetland Classification system varied between each site, with kappa values ranging from 0.64 for the swamp class and 0.89 for the marsh class. This research developed a unique approach to mapping wetlands of various degrees of disturbance using GEOBIA, which can be applied to study other wetlands of similar settings.
Article
Procéder à la restauration de la dynamique naturelle d'un écosystème passe par un suivi scientifique efficace des opérations et de leurs effets écologiques. Après vingt ans de suivi de la réhabilitation de la tourbière de Landemarais en Bretagne, quels sont les résultats sur le maintien des espèces et des habitats ? Quelles sont les améliorations à envisager dans les pratiques de gestion ?
Article
Appropriate integration of remote sensing technologies into ecosystem services concepts and practices leads to potential practical benefits for the protection of biodiversity and the promotion of sustainable use of Earth's natural assets. The last decade has seen the rapid development of research efforts on the topic of ecosystem services, which has led to a significant increase in the number of scientific publications. This systematic review aims to identify, evaluate and synthesise the evidence provided in published peer reviewed studies framing their work in the context of spatially explicit remote sensing assessment and valuation of ecosystem services. Initially, a search through indexed scientific databases found 5920 papers making direct and/or indirect reference to the topic of “ecosystem services” between the years of 1960 and 2013. Among these papers, 211 make direct reference to the use of remote sensing. During the search we aimed at selecting papers that were peer-reviewed publications available through indexed bibliographic databases. For this reason, our literature search did not include books, grey literature, extended abstracts and presentations. We quantitatively present the growth of remote sensing applications in ecosystem services’ research, reviewing the literature to produce a summary of the state of available and feasible remote sensing variables used in the assessment and valuation of ecosystem services. The results provide valuable information on how remotely sensed Earth observation data are used currently to produce spatially-explicit assessments and valuation of ecosystem services. Using examples from the literature we produce a concise summary of what has been done, what can be done and what can be improved upon in the future to integrate remote sensing into ecosystem services research. The reason for doing so is to motivate discussion about methodological challenges, solutions and to encourage an uptake of remote sensing technology and data where it has potential practical applications.
Article
The advancement of synthetic aperture radar (SAR) technology with high-resolution and quad-polarization data demands better and efficient polarimetric SAR (PolSAR) speckle-filtering algorithms. Two requirements on PolSAR speckle filtering are proposed: 1) speckle filtering should be applied to distributed media only, and strong hard targets should be kept unfiltered; and 2) scattering mechanism preservation should be taken into consideration, in addition to speckle reduction. The purpose of this paper is twofold: 1) to propose an effective algorithm that is an extension of the improved sigma filter developed for single-polarization SAR; and 2) to investigate speckle characteristics and the need for speckle filtering for very high resolution (decimeter) PolSAR data. The proposed filter was specifically developed to account for the aforementioned two requirements. Its effectiveness is demonstrated with Jet Propulsion Laboratory airborne synthetic aperture radar data, and comparisons are made with a boxcar filter, the refined Lee filter, and a Wishart-based nonlocal filter. For very high resolution PolSAR systems, such as the German Aerospace Center F-SAR and Japanese Pi-SAR2, with decimeter spatial resolution, we found that the complex Wishart distribution is still valid to describe PolSAR speckle characteristics of distributed media and that speckle filtering may be needed depending on the size of objects to be analyzed. F-SAR X-band data with 25-cm resolution is used for illustration.
Article
Random forests are a combination of tree predictors such that each tree depends on the values of a random vector sampled independently and with the same distribution for all trees in the forest. The generalization error for forests converges a.s. to a limit as the number of trees in the forest becomes large. The generalization error of a forest of tree classifiers depends on the strength of the individual trees in the forest and the correlation between them. Using a random selection of features to split each node yields error rates that compare favorably to Adaboost (Y. Freund & R. Schapire, Machine Learning: Proceedings of the Thirteenth International conference, ***, 148–156), but are more robust with respect to noise. Internal estimates monitor error, strength, and correlation and these are used to show the response to increasing the number of features used in the splitting. Internal estimates are also used to measure variable importance. These ideas are also applicable to regression.
Article
caret has several functions that attempt to streamline the model building and evaluation process. The train function can be used to • evaluate, using resampling, the effect of model tuning parameters on performance • choose the “optimal ” model across these parameters • estimate model performance from a training set To optimize tuning parameters of models, train can be used to fit many predictive models over a grid of parameters and return the “best ” model (based on resampling statistics). See Table 1 for the models currently available. As an example, the multidrug resistance reversal (MDRR) agent data is used to determine a predictive model for the “ability of a compound to reverse a leukemia cell’s resistance to adriamycin” (Svetnik et al, 2003). For each sample (i.e. compound), predictors are calculated that reflect characteristics of the molecular structure. These molecular descriptors are then used to predict assay results that reflect resistance. The data are accessed using data(mdrr). This creates a data frame of predictors called mdrrDescr and a factor vector with the observed class called mdrrClass. To start, we will: • use unsupervised filters to remove predictors with unattractive characteristics (e.g. distributions or high inter–predictor correlations) spare • split the entire data set into a training and test setThe caret Package • center and scale the training and test set using the predictor means and standard deviations from the training set See the package vignette “caret Manual – Data and Functions ” for more details about these operations.> print(ncol(mdrrDescr)) [1] 342> nzv <- nearZeroVar(mdrrDescr)> filteredDescr <- mdrrDescr[,-nzv]> print(ncol(filteredDescr)) [1] 297> descrCor <- cor(filteredDescr)> highlyCorDescr <- findCorrelation(descrCor, cutoff = 0.75)> filteredDescr <- filteredDescr[,-highlyCorDescr]> print(ncol(filteredDescr)) [1] 50> set.seed(1)> inTrain <- sample(seq(along = mdrrClass), length(mdrrClass)/2)> trainDescr <- filteredDescr[inTrain,]> testDescr <- filteredDescr[-inTrain,]> trainMDRR <- mdrrClass[inTrain]> testMDRR <- mdrrClass[-inTrain]> print(length(trainMDRR)) [1] 264> print(length(testMDRR)) [1] 264> preProcValues <- preProcess(trainDescr)> trainDescr <- predict(preProcValues, trainDescr)> testDescr <- predict(preProcValues, testDescr)
Article
Impacts of climate change on US wetlands will add to those of historical impacts due to other causes. In the US, wetland losses and degradation result from drainage for agriculture, filling for urbanization and road construction. States that rely heavily on agriculture (California, Iowa, Illinois, Missouri, Ohio, Indiana) have lost over 80% of their historical area of wetlands, and large cities, such as Los Angeles and New York City, have retained only tiny remnants of wetlands, all of which are highly disturbed. The cumulative effects of historical and future degradation will be difficult to abate. A recent review of mitigation efforts in the US shows a net loss of wetland area and function, even though ‘no net loss’ is the national policy and compensatory measures are mandatory. US policy does not include mitigation of losses due to climate change. Extrapolating from the regulatory experience, one can expect additional losses in wetland areas and in highly valued functions. Coastal wetlands will be hardest hit due to sea-level rise. As wetlands are increasingly inundated, both quantity and quality will decline. Recognition of historical, current and future losses of wetland invokes the precautionary principal: avoid all deliberate loss of coastal wetland area in order to reduce overall net loss. Failing that, our ability to restore and sustain wetlands must be improved substantially.