No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Spatiotemporal Dynamic of Surface Water Bodies Using Landsat Time-Series Data from 1999 to 2011
Abstract
Detailed information on the spatiotemporal dynamic in surface water bodies is important for quantifying the effects of a drying climate, increased water abstraction and rapid urbanization on wetlands. The Swan Coastal Plain (SCP) with over 1500 wetlands is a global biodiversity hotspot located in the southwest of Western Australia, where more than 70% of the wetlands have been lost since European settlement. SCP is located in an area affected by recent climate change that also experiences rapid urban development and ground water abstraction. Landsat TM and ETM+ imagery from 1999 to 2011 has been used to automatically derive a spatially and temporally explicit time-series of surface water body extent on the SCP. A mapping method based on the Landsat data and a decision tree classification algorithm is described. Two generic classifiers were derived for the Landsat 5 and Landsat 7 data. Several landscape metrics were computed to summarize the intra and interannual patterns of surface water dynamic. Top of the atmosphere (TOA) reflectance of band 5 followed by TOA reflectance of bands 4 and 3 were the explanatory variables most important for mapping surface water bodies. Accuracy assessment yielded an overall classification accuracy of 96%, with 89% producer’s accuracy and 93% user’s accuracy of surface water bodies. The number, mean size, and total area of water bodies showed high seasonal variability with highest numbers in winter and lowest numbers in summer. The number of water bodies in winter increased until 2005 after which a decline can be noted. The lowest numbers occurred in 2010 which coincided with one of the years with the lowest rainfall in the area. Understanding the spatiotemporal dynamic of surface water bodies on the SCP constitutes the basis for understanding the effect of rainfall, water abstraction and urban development on water bodies in a spatially explicit way.
... The number of decision trees for the random forest model was set to 501 and the number of randomly selected candidate variables at each split was set to 5. The random forest algorithm is one of the widely used machine learning algorithm to map surface water from satellite images in recent years (e.g. Tulbure and Broich, 2013;Wang et al., 2018;Li and Xu, 2021). Random forest models, trained using the training data from wet and dry seasons to minimise seasonal bias (see Section 3.3), were used to predict the presence of surface water on each pixel for each Landsat image, resulting in an image time series of water presence probability. ...
... Even if an extreme floods occur in the CEB + S, most of the surface water evaporated in few months. Interannual variability in the surface water extent has been observed in other water basins in drylands (Tulbure and Broich, 2013), but the variability in those areas were less strong because of large perennial rivers that transverse such basins. During the thirty two years , the 2008-2011 period was the wettest in the CEB + S resulting in larger surface water extents. ...
... High omission and commission errors for drier years imply that our random forest models were unable to separate correctly water pixels from nonwater pixels during drier years, probably because of high surface reflectivity and shallowness of the water bodies. Our result of low user's and producer's accuracies during dry years is different from the findings of Tulbure and Broich (2013) who also used random forest model to map surface water in Murray-Darling Basin of Australia from Landsat data. Tulbure and Broich (2013) reported lower producer's and user's accuracies of water extent for wet years than dry years. ...
Spatio-temporal characterization of surface water dynamics with Landsat in Cuvelai-Etosha Basin
... The number of decision trees for the random forest model was set to 501 and the number of randomly selected candidate variables at each split was set to 5. The random forest algorithm is one of the widely used machine learning algorithm to map surface water from satellite images in recent years (e.g. Tulbure and Broich, 2013;Wang et al., 2018;Li and Xu, 2021). Random forest models, trained using the training data from wet and dry seasons to minimise seasonal bias (see Section 3.3), were used to predict the presence of surface water on each pixel for each Landsat image, resulting in an image time series of water presence probability. ...
... Even if an extreme floods occur in the CEB + S, most of the surface water evaporated in few months. Interannual variability in the surface water extent has been observed in other water basins in drylands (Tulbure and Broich, 2013), but the variability in those areas were less strong because of large perennial rivers that transverse such basins. During the thirty two years , the 2008-2011 period was the wettest in the CEB + S resulting in larger surface water extents. ...
... High omission and commission errors for drier years imply that our random forest models were unable to separate correctly water pixels from nonwater pixels during drier years, probably because of high surface reflectivity and shallowness of the water bodies. Our result of low user's and producer's accuracies during dry years is different from the findings of Tulbure and Broich (2013) who also used random forest model to map surface water in Murray-Darling Basin of Australia from Landsat data. Tulbure and Broich (2013) reported lower producer's and user's accuracies of water extent for wet years than dry years. ...
Spatio-temporal characterization of surface water dynamics with Landsat in the Cuvelai basin of Namibia
... The number of decision trees for the random forest model was set to 501 and the number of randomly selected candidate variables at each split was set to 5. The random forest algorithm is one of the widely used machine learning algorithm to map surface water from satellite images in recent years (e.g. Tulbure and Broich, 2013;Wang et al., 2018;Li and Xu, 2021). Random forest models, trained using the training data from wet and dry seasons to minimise seasonal bias (see Section 3.3), were used to predict the presence of surface water on each pixel for each Landsat image, resulting in an image time series of water presence probability. ...
... Even if an extreme floods occur in the CEB + S, most of the surface water evaporated in few months. Interannual variability in the surface water extent has been observed in other water basins in drylands (Tulbure and Broich, 2013), but the variability in those areas were less strong because of large perennial rivers that transverse such basins. During the thirty two years , the 2008-2011 period was the wettest in the CEB + S resulting in larger surface water extents. ...
... High omission and commission errors for drier years imply that our random forest models were unable to separate correctly water pixels from nonwater pixels during drier years, probably because of high surface reflectivity and shallowness of the water bodies. Our result of low user's and producer's accuracies during dry years is different from the findings of Tulbure and Broich (2013) who also used random forest model to map surface water in Murray-Darling Basin of Australia from Landsat data. Tulbure and Broich (2013) reported lower producer's and user's accuracies of water extent for wet years than dry years. ...
Socio-economic damages caused by extreme floods have been increasing rapidly in recent years, mainly driven by changes in the climate and modulated by increasing human population in deltic areas and floodplains. The Cuvelai-Etosha Basin (CEB) in southern Africa, covering southern Angola and northern Namibia, experiences socially and economically devastating extreme floods. Yet, accurate information on past and current surface water changes and dynamics is lacking. Here, we estimate and map the surface water extents in the CEB and its surroundings (CEB + S) for 32 years (1990–2021) from Landsat data using random forest models to provide long-term baseline information on surface water changes and dynamics. Based on the reference data, a total of 15,677 ± 1080 km2 have been inundated by surface water in the CEB + S during 1990–2021. This extent was accurately mapped by our local water extent product (mapped area = 16,273 km2, user’s accuracy = 91.5 ± 2.5%, producer’s accuracy = 91.1 ± 6%). With user’s and producer’s accuracy of 91%, our overall water extent provides the first most accurate long-term baseline information on surface water inundation in CEB + S necessary for local spatial planning processes to minimise future negative impacts of floods in the basin. Interannual variability of surface water extent is, however, high, with water extent ranging from 520.8 ± 375.7 km2 to 12372.3 ± 1154.7 km2 during the 1990–2021 period. The largest annual water extents (>10,000 km2) were recorded in 2006, 2008, 2009, 2011, and 2017, whereas the smallest extents (<1000 km2) were recorded in 1992 and 2019. We found that over 40% of the area inundated in the CEB + S during 1990–2021 was inundated less than 9 times. With human population increasing rapidly in the CEB + S, rarely inundated areas with short water residence could become a prime target for human settlements, which may lead to huge socio-economic damages during extreme floods if no preventive measures are put in place. Globally available surface water maps from the Global Land Analysis and Discovery (GLAD) and European Commission’s Joint Research Centre (JRC) did not provide realistic surface water extent for CEB + S, especially during years with extreme floods. Therefore, locally adopted product for operational monitoring of surface water in the CEB + S is needed to provide accurate information for informing spatial planning processes and surface water resource management strategies in this endorheic basin and help minimise future negative impacts of floods.
... Soulard et al. (2022) concluded that complete, cloud-free monthly records spanning multiple decades represent a reasonable temporal resolution for applications focused on intra-and inter-annual surface water dynamics. Recent literature suggests that dense time series are well-suited to capture processes of land surface change at monthly scales (Senay et al., 2017;Tulbure & Broich, 2013;Waylen et al., 2014). ...
... Surface water extent exhibits considerable spatiotemporal variability across the US (Senay et al., 2017;Tulbure & Broich, 2013;Waylen et al., 2014). Spatial stratification is helpful for conveying region-specific dynamics. ...
Satellite imagery is commonly used to map surface water extents over time, but many approaches yield discontinuous records resulting from cloud obstruction or image archive gaps. We applied the Dynamic Surface Water Extent (DSWE) model to downscaled (250-m) daily Moderate Resolution Imaging Spectroradiometer (MODIS) data in Google Earth Engine to generate monthly surface water maps for the conterminous United States (US) from 2003 through 2019. The aggregation of daily observations to monthly maps of maximum water extent produced records with diminished cloud and cloud shadow effects across most of the country. We used the continuous monthly record to analyze spatiotemporal surface water trends stratified within Environmental Protection Agency Ecoregions. Although not all ecoregion trends were significant (p<0.05), results indicate that much of the western and eastern US underwent a decline in surface water over the 17-year period, while many ecoregions in the Great Plains had positive trends. Trends were also generated from monthly streamgage discharge records and compared to surface water trends from the same ecoregion. These approaches agreed on the directionality of trend detected for 54 of 85 ecoregions, particularly across the Great Plains and portions of the western US, whereas trends were not congruent in select western deserts, the Great Lakes region, and the southeastern US. By describing the geographic distribution of surface water over time and comparing these records to instrumented discharge data across the conterminous US, our findings demonstrate the efficacy of using satellite imagery to monitor surface water dynamics and supplement traditional instrumented monitoring.
... Such approaches may prove valuable in identifying waterbodies, which may subsequently be interrogated to determine more detailed wetland characteristics such as class, form, and type. Other decision tree methods have also been used to map wetland classes, flora, and fauna, including: Classification Tree (CT) Analysis [282][283][284][285], Gradient Boosting (GB) [106,[286][287][288], and Classification And Regression Tree (CART) [6,[289][290][291]. Baker et al. [106] noted GB to be preferable to CT approaches for mapping wetland, non-wetland, and riparian land cover classes; however, Tulbure et al. [282] obtained an overall accuracy of 96% when classifying water bodies from other land cover types. ...
... Other decision tree methods have also been used to map wetland classes, flora, and fauna, including: Classification Tree (CT) Analysis [282][283][284][285], Gradient Boosting (GB) [106,[286][287][288], and Classification And Regression Tree (CART) [6,[289][290][291]. Baker et al. [106] noted GB to be preferable to CT approaches for mapping wetland, non-wetland, and riparian land cover classes; however, Tulbure et al. [282] obtained an overall accuracy of 96% when classifying water bodies from other land cover types. Pantaleoni et al. [289] noted that CART better classified three wetland classes from upland land cover types with 73% overall accuracy. ...
The Prairie Pothole Region (PPR) of North America is an extremely important habitat for
a diverse range of wetland ecosystems that provide a wealth of socio-economic value. This paper describes the ecological characteristics and importance of PPR wetlands and the use of remote sensing for mapping and monitoring applications. While there are comprehensive reviews for wetland remote sensing in recent publications, there is no comprehensive review about the use of remote sensing in the PPR. First, the PPR is described, including the wetland classification systems that have been used, the water regimes that control the surface water and water levels, and the soil and vegetation characteristics of the region. The tools and techniques that have been used in the PPR for analyses of geospatial data for wetland applications are described. Field observations for ground truth data are critical for good validation and accuracy assessment of the many products that are produced. Wetland classification approaches are reviewed, including Decision Trees, Machine Learning, and object versus pixel-based approaches. A comprehensive description of the remote sensing systems and data that have been employed by various studies in the PPR is provided. A wide range of data can be used for various applications, including passive optical data like aerial photographs or satellite-based, Earth-observation data. Both airborne and spaceborne lidar studies are described.
A detailed description of Synthetic Aperture RADAR (SAR) data and research are provided. The state of the art is the use of multi-source data to achieve higher accuracies and hybrid approaches. Digital Surface Models are also being incorporated in geospatial analyses to separate forest and shrub and emergent systems based on vegetation height. Remote sensing provides a cost-effective mechanism for mapping and monitoring PPR wetlands, especially with the logistical difficulties and cost of field-based methods. The wetland characteristics of the PPR dictate the need for high resolution in both time and space, which is increasingly possible with the numerous and increasing remote sensing
systems available and the trend to open-source data and tools. The fusion of multi-source remote sensing data via state-of-the-art machine learning is recommended for wetland applications in the PPR. The use of such data promotes flexibility for sensor addition, subtraction, or substitution as a function of application needs and potential cost restrictions. This is important in the PPR because of the challenges related to the highly dynamic nature of this unique region.
... STIS peut être vu comme une pile d'images (une image par date) ou comme une grille de séries temporelles où chaque pixel est associé à sa série temporelle. Les STIS ont été utilisés dans de nombreuses applications d'utilisation des terres / de couverture terrestre, par exemple, pour surveiller la dynamique de la végétation [161] ou les plans d'eau [163], détecter les changements de couverture terrestre [193], analyser les effets à long terme du changement climatique [64], suivi des dommages causés par une catastrophe [173]. Récemment, plusieurs méthodes ont été proposées pour analyser les STIS telles que la déformation temporelle dynamique [129], le cube de données spatio-temporelles [183], l'apprentissage en profondeur [121] ou les modèles séquentiels [98] etc. Cependant, il existe encore un vaste domaine de recherche pour surmonter les défis posés par STIS et trouver des méthodes efficaces pour les utiliser dans des applications du monde réel. ...
... SITS can be seen as a stack of images (one image per date) or as a grid of time series where each pixel is associated with its time series. SITS have been used in many land use/land cover applications, e.g., monitoring vegetation dynamics [161] or water bodies [163], de-tecting land cover changes [193], analyzing long-term effects of climate change [64], tracking disaster damages [173]. Recently, several methods have been proposed to analyze SITS such as dynamic time warping [129], spatio-temporal data cube [183], deep learning [121], or sequential patterns [98] etc. ...
Although morphological hierarchies are today a well-established framework for single frame image processing, their extension to time-related data remains largely unexplored. This thesis aims to tackle the analysis of satellite image time series with tree-based representations. To do so, we distinguish between three kinds of models, namely spatial, temporal and spatio-temporal hierarchies. For each model, we propose a streaming algorithm to update the tree when new images are appended to the series. Besides, we analyze the structural properties of the different tree building strategies, thus requiring some projection methods for the spatio-temporal tree in order to obtain comparable structures. Then, trees are compared according to their node distribution, filtering capability and cost, leading to a superiority of the spatio-temporal tree (a.k.a. space-time tree). Hence, we review spatio-temporal attributes, including some new ones, that can been extracted from the space-time tree in order to compute some multiscale features at the pixel or image level. These attributes are finally involved in tools such as filtering and pattern spectrum for various remote sensing based applications.
... Water bodies have been mapped and detected by optical and radar imaging for decades (Van Dijk and Renzullo, 2011). Imageries taken from remote sensing devices have been widely used to detect changes occurred on water level of lakes or any other Earth surface feature (Bagli et al., 2004;Li et al., 2010;Tulbure and Broich, 2013;Wang et al., 2013;Song et al., 2014;Jawak et al., 2015;Zhai et al., 2015;Sisay, 2016;Zhang et al., 2017;Huang et al., 2018). They are commonly used for mapping and monitoring lakes and reservoirs, water quality mapping and modeling, drainage network mapping and hydrological research, basin characterization and pollution researches on surface water and ground water. ...
... Many researchers have applied the Landsat series of images to extract features of surface water and analyze space-time changes in different study areas (e.g. El Gammal et al., 2010;Dinka, 2012;Tulbure and Broich, 2013;Rokni et al., 2014;Mishra and Prasad, 2015;Tulbure et al., 2016;Sarp and Ozcelik, 2017;Mustafa and Bayat, 2019;Carstens and Amer, 2019) Several methods have been proposed to extract water bodies from Landsat images. The common water mapping methods are categorized as: spectral analysis, single-band thresholding method, supervised and unsupervised classification and spectral water index method (Zhang et al., 2003). ...
Changes in lake surface area and fluctuations in water levels are common, especially in the Ethiopian Rift Valley region. The aim of this study was to present the space-time changes of lake Beseka from 1985 to 2020 using multitemporal Landsat images. Four spectral water indices, namely Normalized Differential Water Index (NDWI), Modified Normalized Difference Water Index (MNDWI) and Automated Water Extraction Index (AWEI) were utilized for the extraction of the lake water. The results of the study show that lake Beseka has tended to expand significantly and strongly over the past 35 years. During 1985 the lake surface area was around 30.9 km2 and within the next 35 years, the lake surface area increased by 18.18 km2and in 2020 its is about 49.1 km2. Lake Beseka shows 0.2412km2 area change from 2015 to 2020. This indicates Beseka has a decreasing trend of expansion compared to its history. However; The topography of the lake is lower in elevation, a small water level rise can cover a large area and lead to significant impacts on the surrounding environment. Therefore, monitoring of lake surface changes by multi-temporal satellite images is very necessary and of decisive importance.
... Remote sensing models are often validated using independent and higher resolution imagery; however, using field data for validation was more powerful and allowed us to test the sensitivity and limitations of the remotely sensed data (Alonso et al., 2020;Foody, 2002). We made use of pre-existing field transect data instead of using remote sensing data for validation, as is common in other studies (Alonso et al., 2020;Thomas et al., 2011;Tulbure & Broich, 2013). The challenge we faced with making use of the field data to validate the remote sensing model was the different spatial resolutions of the validation plots (5 m) and Landsat pixels (30 m). ...
... Sampling is sometimes done in a way that all validation points are contained within a large area of the relevant ground cover away from boundaries (Jahncke et al., 2018), which leads to a potential bias towards a higher accuracy estimate (Foody, 2002). In our study, transects were placed on or near the wet/dry boundary where more classification errors were likely; however, our overall accuracy and kappa statistic were both high, and comparable to similar studies mapping inundation of wetlands within Australia using Landsat data suggesting that our method was sufficient for validation (Thomas et al., 2011;Tulbure & Broich, 2013). ...
Understanding broad trends in the distribution and composition of wetlands is essential for making evidence-based management decisions. Determining temporal change in the extent of inundation in wetlands using remote sensing remains challenging and requires on-ground verification to determine accuracy and precision. Therefore, optimization and validation of remote sensing methods in threatened wetlands is a high priority for their conservation. Despite their ecological importance in the landscape, we have little knowledge of the variation in the spatial extent of inundation in upland lagoons, a threatened ecological community in New South Wales, Australia. Our project developed locally trained algorithms to predict the extent of water and emergent vegetation using imagery from the Landsat-5,-7, and-8 satellites. The best model for upland lagoons used shortwave infrared reflectance (performing better than normalized difference spectral indices), with model accuracy against validation transects greater than 95%. We applied the model to images from 1988 to 2020 across 58 lagoons to generate a dataset that demonstrates the variable water regime and vegetation change in response to local rainfall over 32 years such as in the lagoons. Our results reduce threats to a dynamic threatened ecological community by filling an important knowledge gap and demonstrate a valuable method to understand historical and current changes in the hydrology of dynamic wetland systems more broadly.
... In drylands, the occurrence of water is intermittent and difficult to predict, and earth observations represent the only way of quantifying and monitoring these dynamics across large spatial extents and through time (Leblanc et al., 2012). The availability of public domain, temporally deep satellite data archives, together with improvements in algorithms and available computing power , have led to huge leaps in recent years in mapping surface water dynamics at the regional (Carroll et al., 2017;Tulbure and Broich, 2013;Tulbure et al., 2016;Zou et al., 2018), continental (Jones, 2019(Jones, , 2015, and global scales (Pekel et al., 2016;Pickens et al., 2020). Most of these mapping efforts have been based on optical satellite data such as MODIS (Klein et al., 2017) at daily intervals but coarser resolution (250 m) or the entire Landsat archive at low temporal resolution (16-day repeat frequency) and medium spatial resolution (30 m). ...
... Our overall accuracy was 99% with user's accuracy (complement of commission error) for the water class of 80% (±3.6%) and producer's accuracy (complement of omission error) for the water class of 76% (±5.6%), similar to other studies mapping surface water/wetlands (Bwangoy et al., 2010;Midekisa et al., 2014;Tulbure et al., 2016;Tulbure and Broich, 2013;Wright and Gallant, 2007) and to what is expected of remote sensing accuracy assessments (Foody, 2008). In addition, our work here mapped not only open water, but also floods, which are more challenging to detect than open water because their spectral signature can be different from that of permanent water or seasonally inundated areas due to sediment load, turbidity, dissolved matter, algal content, depth, and bottom reflectance signal (Pekel et al., 2016). ...
Spatiotemporal quantification of surface water and flooding is essential given that floods are among the largest natural hazards. Effective disaster response management requires near real-time information on flood extent. Satellite remote sensing is the only way of monitoring these dynamics across vast areas and over time. Previous water and flood mapping efforts have relied on optical time series, despite cloud contamination. This reliance on optical data is due to the availability of systematically acquired and easily accessible optical data globally for over 40 years. Prior research used either MODIS or Landsat data, trading either high temporal density but lower spatial resolution or lower cadence but higher spatial resolution. Both MODIS and Landsat pose limitations as Landsat can miss ephemeral floods, whereas MODIS misses small floods and inaccurately delineates flood edges. Leveraging high temporal frequency of 3–4 days of the existing Landsat-8 (L8) and two Sentinel-2 (S2) satellites combined, in this research, we assessed whether the increased temporal frequency of the three sensors improves our ability to detect surface water and flooding extent compared to a single sensor (L8 alone). Our study area was Australia’s Murray-Darling Basin, one of the world’s largest dryland basins that experiences ephemeral floods. We applied machine learning to NASA’s Harmonized Landsat Sentinel-2 (HLS) Surface Reflectance Product, which combines L8 and S2 observations, to map surface water and flooding dynamics. Our overall accuracy, estimated from a stratified random sample, was 99%. Our user’s and producer’s accuracy for the water class was 80% (±3.6%, standard error) and 76% (±5.8%). We focused on 2019, one of the most recent years when all three HLS sensors operated at full capacity. Our results show that water area (permanent and flooding) identified with the HLS was greater than that identified by L8, and some short-lived flooding events were detected only by the HLS. Comparison with high resolution (3 m) PlanetScope data identified extensive mixed pixels at the 30 m HLS resolution, highlighting the need for improved spatial resolution in future work. The HLS has been able to detect floods in cases when one sensor (L8) alone was not, despite 2019 being one of the driest years in the area, with few flooding events. The dense optical time-series offered by the HLS data is thus critical for capturing temporally dynamic phenomena (i.e., ephemeral floods in drylands), highlighting the importance of harmonized data such as the HLS.
... As an important part of the land-water cycle, surface water resources play an important role in promoting national economic development and maintaining the balance of terrestrial and aquatic ecosystems, agriculture, and the ecological environment [1]. In terms of global climate change, all kinds of water resources show obvious characteristics of temporal and spatial differentiation [2][3][4][5]. It is necessary to strengthen the dynamic monitoring and investigation of water resources, especially for arid and semi-arid areas. ...
The spatio-temporal change of the surface water is very important to agricultural, economic, and social development in the Hetao Plain, as well as the structure and function of the ecosystem. To understand the long-term changes of the surface water area in the Hetao Plain, we used all available Landsat images (7534 scenes) and adopted the modified Normalized Difference Water Index (mNDWI), Enhanced Vegetation Index (EVI), and Normalized Difference Vegetation Index (NDVI) to map the open-surface water from 1989 to 2019 in the Google Earth Engine (GEE) cloud platform. We further analyzed precipitation, temperature, and irrigated area, revealing the impact of climate change and human activities on long-term surface water changes. The results show the following. (1) In the last 31 years, the maximum, seasonal, and annual average water body area values in the Hetao Plain have exhibited a downward trend. Meanwhile, the number of maximum, seasonal, and permanent water bodies displayed a significant upward trend. (2) The variation of the surface water area in the Hetao Plain is mainly affected by the maximum water body area, while the variation of the water body number is mainly affected by the number of minimum water bodies. (3) Precipitation has statistically significant positive effects on the water body area and water body number, which has statistically significant negative effects with temperature and irrigation. The findings of this study can be used to help the policy-makers and farmers understand changing water resources and its driving mechanism and provide a reference for water resources management, agricultural irrigation, and ecological protection.
... Intensified climate variations can severely affect surface water and cause strong seasonal dynamics in surface water [3]. Several water bodies have experienced both seasonal and interannual variations over the years [20,[25][26][27][28]. Previous studies have assessed multidecadal changes of water area either by using one or a few Landsat scenes per year [29] or standalone years [30]. ...
Sri Lanka contains a large number of natural and man-made water bodies, which play an essential role in irrigation and domestic use. The island has recently been identified as a global hotspot of climate change extremes. However, the extent, spatial distribution, and the impact of climate and anthropogenic activities on these water bodies have remained unknown. We investigated the distribution, spatial and temporal changes, and the impacts of climatic and anthropogenic drivers on water dynamics in Dry, Intermediate, and Wet zones of the island. We used Landsat 5 and Landsat 8 images to generate per-pixel seasonal and annual water occurrence frequency maps for the period of 1988-2019. The results of the study demonstrated high inter-and intra-annual variations in water with a rapid increase. Further, results showed strong zonal differences in water dynamics, with most dramatic variations in the Dry zone. Our results revealed that 1607.73 km 2 of the land area of the island is covered by water bodies, among this 882.01 km 2 (54.86%) is permanent and 725.72 km 2 (45.14%) is seasonal water area. Total inland seasonal water increased with a dramatic annual growth rate of 7.06 ± 1.97 km 2 compared to that of permanent water (4.47 ± 2.08 km 2 /year). Sri Lanka has the highest permanent water area during December-February (1045.97 km 2), and drops to the lowest in May-September (761.92 km 2) when the seasonal water (846.46 km 2) is higher than permanent water. The surface water area was positively related to both precipitation and Gross Domestic Product, while negatively related to the temperature. Findings of our study provide important insights into possible spatiotemporal changes in surface water availability in Sri Lanka under certain climate change and anthropogenic activities.
... Hydrologic remote sensing focused on high-altitude or satellite-based products. Traditional ground-based and hydraulically-modeled discharges have been coupled with various satellite platforms to compute discharge using remote sensing and include (1) radar altimetry from National Aeronautics and Space Administration (NASA) satellites TOPEX/Poseidon, ERS-1, 2; ENVISAT; NASA SRTM; NASA Jason-2; SARAL; CryoSat-2; and Sentinel 3A [15][16][17][18][19][20][21][22][23][24][25][26][27]; (2) NASA Landsat imagery [28][29][30][31][32][33][34][35][36]; (3) AMSR-E [37], and (4) NASA SWOT [1,[38][39][40][41][42]. For example, satellitebased discharges were derived for the Yukon River at Eagle and Stevens Point, Alaska, based on water-surface area, slope, and water-surface height [25]. ...
The U.S. Geological Survey is actively investigating remote sensing of surface velocity and river discharge (discharge) from satellite-, high altitude-, small, unmanned aircraft systems- (sUAS or drone), and permanent (fixed) deployments. This initiative is important in ungaged basins and river reaches that lack the infrastructure to deploy conventional streamgaging equipment. By coupling alternative discharge algorithms with sensors capable of measuring surface velocity, streamgage networks can be established in regions where data collection was previously impractical or impossible. To differentiate from satellite or high-altitude platforms, near-field remote sensing is conducted from sUAS or fixed platforms. QCam is a Doppler (velocity) radar mounted and integrated on a 3DR© Solo sUAS. It measures the along-track surface velocity by spot dwelling in a river cross section at a vertical where the maximum surface velocity is recorded. The surface velocity is translated to a mean-channel (mean) velocity using the probability concept (PC), and discharge is computed using the PC-derived mean velocity and cross-sectional area. Factors including surface-scatterer quality, flight altitude, propwash, wind drift, and sample duration may affect the radar-returns and the subsequent computation of mean velocity and river discharge. To evaluate the extensibility of the method, five science flights were conducted on four rivers of varying size and dynamics and included the Arkansas River, Colorado (CO), USA (two events); Salcha River near Salchaket, Alaska (AK), USA; South Platte River, CO, USA; and the Tanana River, AK, USA. QCam surface velocities and river discharges were compared to conventional streamgaging methods, which represented truth. QCam surface velocities for the Arkansas River, Salcha River, South Platte River, and Tanana River were 1.02 meters per second (m/s) and 1.43 m/s; 1.58 m/s; 0.90 m/s; and 2.17 m/s, respectively. QCam discharges (and percent differences) were 9.48 (0.3%) and 20.3 cubic meters per second (m3/s) (2.5%); 62.1 m3/s (−10.4%); 3.42 m3/s (7.3%), and 1579 m3/s (−18.8%). QCam results compare favorably with conventional streamgaging and are a viable near-field remote sensing technology that can be operationalized to deliver real-time surface velocity, mean velocity, and river discharge, if cross-sectional area is available.
... The second group of methods are statistical methods, such as those using multivariate regression or discriminant analysis. Classification methods (the third group) are a matrix of combinations of different methods-this is a pixel or objectoriented approach, classifications with or without training, various classification machines; for example: a random forest or support vector machine, neural algorithms [28][29][30][31][32]. There are also various special techniques (group four) such as entropy-based computer vision techniques [33]. ...
In semi-arid ecoregions of temperate zones, focused snowmelt water infiltration in topographic depressions is a key, but imperfectly understood, groundwater recharge mechanism. Routine monitoring is precluded by the abundance of depressions. We have used remote-sensing data to construct mass balances and estimate volumes of temporary ponds in the Tambov area of Russia. First, small water bodies were automatically recognized in each of a time series of high-resolution Planet Labs images taken in April and May 2021 by object-oriented supervised classification. A training set of water pixels defined in one of the latest images using a small unmanned aerial vehicle enabled high-confidence predictions of water pixels in the earlier images (Cohen’s Κ = 0.99). A digital elevation model was used to estimate the ponds’ water volumes, which decreased with time following a negative exponential equation. The power of the exponent did not systematically depend on the pond size. With adjustment for estimates of daily Penman evaporation, function-based interpolation of the water bodies’ areas and volumes allowed calculation of daily infiltration into the depression beds. The infiltration was maximal (5–40 mm/day) at onset of spring and decreased with time during the study period. Use of the spatially variable infiltration rates improved steady-state shallow groundwater simulations.
... Issues arising from cloud contaminated data are magnified in studies that use large numbers of images. Since the Landsat archive opened in 2008, granting free access to historical imagery (Woodcock et al., 2008), time series analyses have proliferated across a variety of applications, including land cover and land use classification (Franklin et al., 2015;Zhu and Woodcock, 2014b), disturbance monitoring (Huang et al., 2010;Kennedy et al., 2007;Zhu et al., 2020) and surface water mapping (Pekel et al., 2016;Tulbure and Broich, 2013). Multi-temporal studies often employ all useable images, a process which relies on first identifying cloud and cloud shadow-free observations. ...
Cloud Cloud shadow Fmask Tmask MAJA Sen2Cor LaSRC Cloud detection Cloud mask A B S T R A C T Accurate, automated cloud and cloud shadow detection is a key component of the processing needed to prepare optical satellite imagery for scientific analysis. Many existing cloud detection algorithms rely on temperature information to identify clouds, making detection difficult for imagers that lack a thermal band, like Sentinel-2. To get maximum benefit from Sentinel-2 products it is critical to understand which algorithms best identify clouds and their shadows in images. We examined the relative performance of five different cloud-masking algorithms (Sen2Cor, MAJA, LaSRC, Fmask and Tmask) in 6 Sentinel-2 scenes (28 total images) distributed across the Eastern Hemisphere. Expanding on these comparisons, we tested ensemble approaches to improve results. We tested three ensemble approaches to cloud and shadow classification based on the outputs of the five initial algorithms using the cloud masks in: (1) a majority prediction model; (2) a random forests model; and (3) a conditional logic model. Accuracy assessments show a trade-off between omission and commission errors in cloud detection for individual algorithms across all sites, and some algorithms are better at detecting either clouds or cloud shadows. No single algorithm outperforms the others for both clouds and shadows. Aggregating the results from multiple algorithms produces fewer undetected clouds and higher overall accuracy than any single algorithm, with as high as 2.7% improvement over the top-performing algorithm, suggesting an ensemble approach may be the most useful for processing of Sentinel-2 data.
... To do this, we use time series of Landsat data from 1984 to 2019 to study the 10 high Andean wetlands detailed in Fig. 1. The Landsat satellite products provide reliable data for studying wetlands and are thus extensively used for characterizing long-term temporal changes in water and vegetation areas (Bortels et al., 2011;Kayastha et al., 2012;Tulbure and Broich, 2013;Banerjee et al., 2016;Mueller et al., 2016;Bowen et al., 2017). Furthermore, to characterize the climate variability in these systems, we study the long-term relation between the changes in water and vegetation areas with rainfall and evaporation variability. ...
High Andean wetlands of the elevated plateaus of the Andes Mountains of Chile, Argentina, Perú and Bolivia are true oases that sustain life in this arid region. Despite their ecological value, they have been rarely studied and are vulnerable to climate change and human activities that require groundwater resources. One such activity that may be intensified in the near future is mining for nonmetallic minerals such as lithium, whose worldwide demand is expected to increase with the rise of electric vehicles that need batteries. To determine a baseline of the natural dynamics of these systems, which allows sustainable management, it is essential to understand the spatiotemporal dynamics of these wetlands. In this article, we studied the temporal and spatial dynamics of high Andean wetlands of Chile, with the aim of identifying the key processes that govern their dynamics. To do this, we used time series of Landsat data from 1984 to 2019 to study 10 high Andean wetlands. Furthermore, to characterize the climate variability in these systems, we studied the long-term relation between the changes in water and vegetation areas with rainfall and evaporation variability. It was found that the groundwater reservoir plays a key role in sustaining the high Andean wetlands. Wet years with a period of occurrence of 20-30 years are the years in which the groundwater reservoirs are actually recharged, and in between wet years, the groundwater reservoirs gradually release the water that sustains the aquatic ecosystems. Hence, groundwater exploitation should be carefully designed from a long-term perspective, as groundwater levels could take decades to recover.
... While direct water surface elevation measurements may not be available for most water bodies due to the lack of in-situ monitoring or altimetry-based measurements, water surface area, assuming accurate detection methods are employed, is relatively easy to obtain and can be directly measured using optical and synthetic aperture radar imagery. Water surface area measurements are also scalable to regional and global levels (Pekel et al., 2016;Tulbure & Broich, 2013;Tulbure et al., 2016). While useful in a wide range of applications, water surface area remains a two-dimensional measurement whereas, at minimum, water volume change estimates are needed to improve our understanding of surface water dynamics and its effects upon climate and other fields. ...
Multidecadal inland surface water dynamics are of increasing interest due to their importance to climate, ecology, and society, yet several key challenges impede long‐term monitoring of inland surface waters globally. This research investigates two novel methods, one addressing subhydroflattened surface estimate uncertainty issues, and a second addressing temporal resolution issues, using 46 water bodies across the western United States. First, low water level estimate uncertainty was reduced using multiple digital elevation models (ALOS, SRTM, and NED) to derive the hypsometric relationship for each lake from the digital elevation model with the lowest hydroflattened water surface. This technique reduced the number of images with subhydroflattened water surfaces by at least 549 over the best individual DEM resulting in higher water surface elevation estimate accuracy. Second, this paper introduces proportional hypsometry which dynamically generates surface area/elevation relationships for every image using clear pixels only by removing contamination from both the image and DEM. Proportional hypsometry was found to be ill‐suited for subhydroflattened water surface levels but produced comparable accuracy to clear images for above hydroflattened water levels. Overall, using the lowest hydroflattened surface along with proportional hypsometry improved temporal resolution enabling analysis of nearly 10,000 additional images while maintaining similar accuracy levels as images with <1% contamination (2.35 m RMSE vs. 2.17 m RMSE). This research increases lower water elevation estimate accuracy and temporal resolution and is scalable enabling regional and global water dynamic analysis.
... DEA provides free and open earth observation data from two public-good satellite programs: Landsat [47] and Sentinel 2 [48]. The Landsat program, operated by the United States Geological Survey and National Aeronautics and Space Administration (NASA), consists of three multi-spectral satellites/sensors: Landsat 5 TM (1984-2013, Landsat 7 ETM+ (1999-present), and Landsat 8 OLI (2013-present). These sensors have a 30 m pixel resolution, and image the Earth every 16 days on average. ...
Water detection algorithms are now being routinely applied to continental and global archives of satellite imagery. However, water resource management decisions typically take place at the waterbody rather than pixel scale. Here, we present a workflow for generating polygons of persistent waterbodies from Landsat observations, enabling improved monitoring and management of water assets across Australia. We use Digital Earth Australia’s (DEA) Water Observations from Space (WOfS) product, which provides a water classified output for every available Landsat scene, to determine the spatial locations and extents of waterbodies across Australia. We generated a polygon set of waterbodies that identified 295,906 waterbodies ranging in size from 3125 m2 to 4820 km2. Each polygon was used to generate a time series of WOfS, providing a history of the change in surface area of each waterbody every ~16 days since 1987. We demonstrate the applications of this new dataset, DEA Waterbodies, to understanding local through to national-scale surface water spatio-temporal dynamics. DEA Waterbodies provides new insights into Australia’s water availability and enables the monitoring of important landscape features such as lakes and dams, improving our ability to use earth observation data to make meaningful decisions.
... Etudes Spatiales, CNES) 主持的地球观测卫星系统 1-5 (SPOT 1-5) [19][20] 等； 由欧洲空间局主持的哨兵系列 地球观测任务 (Sentinel-2) [21] 。③ 高分辨率传感器， 如 IKONOS [22] 、 RapidEye [23] 、Worldview [24] 、ZY-3 [25] 、 Quickbird [26] 、 GF-1/2 [27] 和 SPOT-6 等数据。其中 Sen-tinel-2(S2)数据自 2015 年起可获取， 包含 13 个波 段， 其中 6 个与 Landsat TM 和 ETM+相似 [28] ， 属于新 兴可免费获取的高时空分辨率影像， 为了制定该数 据的最佳算法和处理流程， 科学界向系统开发者的 反馈尤为重要， 目前已开展很多相关研究包括 S2 影 像去云 [29] 及大气顶层反射率数据向地表反射率数 转换等问题。 从光学卫星影像中提取未被植被遮挡的水体 时， 主要利用了水体对近红外以及短波红外电磁辐 射的强吸收特性， 该方法已较成熟 [30] 。其具体应用 又分为单波段法和指数法。其中单波段法是指通 过分析单个近红外波段而得出水体范围 [31] ； 指数法 被证明可以有效的提取水体， 常用的水体指数有归 一化植被指数(Normalized Difference Water Index, NDWI) [32] 、 改进的归一化水体指数(Modified Normalized Difference Water Index, MNDWI) [33] 、 自动水 体 提 取 指 数(Automated Water Extraction Index, AWEI) [34] 等。相比于单一分类方法， 决策树 [35][36][37] ...
Bosten Lake is a typical inland lake in the arid zone. The change in the lake area is strongly related to local natural and cultural environmental changes. Based on the GIS and RS technologies, this paper combines Landsat imagery and MODIS data, including a total of 2289 scenes, with JRC GSW water mask products to characterize the interannual and intraannual changes of the area of Bosten Lake from 2000 to 2019 through the Google Earth Engine (GEE) platform using index methods. We use the 2019 Sentinel-2 images to compare and analyze the results. To quantify the the causes of the changes, we analyzed the human activities and daily meteorological data of Yanqi, Korla and Bayanbuluk meteorological stations during 2000-2018. Results show that: (1) the GEE is efficient for integrating multi-temporal high-resolution remote sensing data to analyze the temporal change of lake area, especially the intraannual change. Compared with Landsat-5/7/8 and MOD09GQ data, the lake shoreline extracted based on Sentinel-2 images shows more details due to their high temporal and spatial resolution; (2) during 2000-2013, the total lake area decreases by 181.66 km 2 with a decreasing rate of 13.98km 2 /a; while during 2013-2019, the lake area increases by 133.13 km 2 with a increasing rate of 22.19 km 2
... In fact, many approaches have been developed for change detection using medium-or high-density stack of NOAA-AVHRR, MODIS, and Landsat imagery over time (Kennedy et al. 2010;Verbesselt et al. 2010b;Zhu and Woodcock 2014). Most of these approaches were developed for and successfully applied to vegetation changes in forest and agricultural landscapes (Kennedy et al. 2007;DeVries et al. 2015;Yin et al. 2018), or to changes in water bodies, tidal flats, and wetlands (Tulbure and Broich 2013;Bishop-Taylor et al. 2018;Wang et al. 2018). Although time-series data have been increasingly used in urban landscapes (Li et al. , 2018bSong et al. 2016;Zhu et al. 2016;Huang et al. 2017;Lu et al. 2017;Liu et al. 2020), when compared to the extensive and successful use in forest, agriculture and wetland areas, its urban use remains challenging for three reasons. ...
ContextUrban landscapes are highly dynamic with changes frequently occurring at short time intervals. Although the Landsat data archive allows the use of high-density time-series data to quantify such dynamics, the approaches that can fully address the spatial and temporal complexity of the urban landscape are still lacking.ObjectivesA new approach is presented for accurately quantifying urban landscape dynamics. Information regarding when and where a change occurs, what type of change exists, and how often it happens are incorporated.Methods
The new approach integrates object-based image analysis and time-series change detection techniques by using all available Landsat images for several decades. This approach was tested on the rapidly urbanizing city of Shenzhen, China from 1986 to 2017.ResultsLand cover changes in both long- and short-time intervals can be proficiently detected with an overall accuracy of 90.65% and a user’s accuracy of 92.18% and 82.40% for “No change” and “Change”, respectively. The frequency and time of change can be explicitly displayed while incorporating the advantages of object-based image analysis and time-series change detection. The efficiency of the change analysis can be greatly increased because the object-based analysis greatly reduces the number of analyzed units.Conclusion
The new approach can accurately and efficiently detect the land cover change for quantifying urban landscape dynamics. Integrating the object and the remotely sensed time-series data has the potential to link the physical and socio-economic properties together for facilitating sustainable landscape planning.
... Landsat satellites (MSS, TM, ETM +, and OLI) that continuously provide medium resolution images are among the most widely used optical sensors in environmental research in the last five decades. The use of Landsat satellite images has an important place in numerous studies where the water surface areas of wetlands are extracted or the temporal changes are determined [24][25][26][27][28][29][30][31][32][33][34][35][36][37][38][39][40][41][42][43]. ...
Ramsar Convention (RC) is the first of modern intergovernmental agreement on the conscious use and conservation of natural resources. It provides a platform for contracting parties working together to develop the best available data, advice, and policy recommendations to increase awareness of the benefits of wetlands in nature and society. Turkey became a party of the RC in 1994, and in the years 1994 to 2013, 14 wetlands that reached the Ramsar criteria were recognized as Ramsar sites (RS). With this study, all inland RS in Turkey from 1985 to 2020 were examined, and changes in the water surface areas were evaluated on the GEE cloud computing platform using Landsat satellite images and the NDWI index. The closest meteorological station data to each RS were evaluated and associated with the surface area changes. The reasons for the changes in these areas, besides the meteorological effects, have been scrutinized using management plans and publications. As a result, inland wetlands decreased at different rates from 1985 to 2020, with a total loss of 31.38% and 21571.0 ha for the spring months. Since the designation dates of RS, the total amount of water surface area reduction was 27.35 %, constituting 17,758.90 ha.
... Verpoorter et al. (2014) used annual composited images to produce the Global Water Bodies database (GLOWABO), comprising all lakes > 0.002 km. Additionally, Heimhuber et al. (2016), Mueller et al. (2016), Sheng et al. (2016), Schaffer-Smith et al. (2017), and Tulbure and Broich (2013) selected multi-temporal cloudless images to remotely model surface water dynamics at a regional scale. Furthermore, Pekel et al. (2016) developed a dataset of global water surface maps using Landsat 5, 7, and 8 data between 1984 and 2015, revealing variations in surface water at a monthly time-scale. ...
Optical remote sensing imagery is commonly used to monitor the spatial and temporal distribution patterns of inland waters. Its usage, however, is limited by cloud contamination, which results in low-quality images or missing values. Selecting cloud-free scenes or combining multi-temporal images to produce a cloud-free composite image can partially overcome this problem at the cost of the monitoring frequency. Predicting the spectral values of cloudy areas based on the spectral characteristics is a possible solution; however, this is not appropriate for water because it changes rapidly. Reconstructing cloud-covered water areas using historical water-distribution data has good performance, but such methods are typically only suitable for lakes and reservoirs, not over vast and complex terrain. This paper proposes a category-based approach to reconstruct the water distribution in cloud-contaminated images using a spatiotemporal dependence model. The proposed method predicts the class label (water or land) of a cloudy pixel based on the neighboring pixel labels and those at the same position in images acquired on other dates according to historical spatiotemporal water-distribution data. The method was evaluated through eight experiments in different study regions using Landsat and Sentinel-2 images. The results demonstrated that the proposed method could yield high-quality cloud-free classification maps and provide good water-extraction accuracy and consistency in most hydrological conditions, with an overall accuracy of up to 98%. The accuracy and practicality of the method render it promising for applications across a wide range of future research and monitoring efforts.
... Presently, a variety of remotely sensed image sources are available for surface water mapping (Alsdorf et al., 2007;Huang et al., 2018aHuang et al., , 2018bPekel et al., 2016;Seaton et al., 2020;Tulbure and Broich, 2013). One of the most popular sources of remotely sensed imagery for reservoir monitoring is the series of sensors that have been carried by the Landsat satellites (Arvor et al., 2018;Avisse et al., 2017;Duan and Bastiaanssen, 2013;Ma et al., 2019;Sheng et al., 2016;Yao et al., 2019;Zhang et al., 2017;Zhao and Gao, 2018). ...
Information on the temporal variation of surface water area of reservoirs is fundamental for water resource management and is often monitored by satellite remote sensing. Moderate Resolution Imaging Spectroradiometer (MODIS) imagery is an attractive data source for the routine monitoring of reservoirs, however, the accuracy is often limited due to the negative impacts associated with its coarse spatial resolution and the effects of cloud contamination. Methods have been proposed to solve these two problems independently but it remains challenging to address both problems simultaneously. To overcome this, this paper proposes a new approach that aims to monitor reservoir surface water area variations accurately and timely from daily MODIS images by exploring sub-pixel scale information. The proposed approach used estimates of reservoir water areas obtained from cloud-free and relatively fine spatial resolution Landsat images and water fraction images by spectral unmixing of coarse MODIS imagery as reference data. For each MODIS pixel, these reference reservoir water areas and their corresponding pixel water fractions were used to construct a linear regression equation, which in turn may be applied to predict the time series of reservoir water areas from daily MODIS water fraction images. The proposed approach was assessed with 21 reservoirs, where the correlation coefficients between reservoir water areas predicted by the common pixel-based analysis method and altimetry water levels were all less than 0.5. With the proposed sub-pixel analysis method, the resultant correlation coefficients were much improved, with eleven values larger than 0.5 including six values larger than 0.8 and the highest value of 0.94. The results show that the proposed sub-pixel analysis method is superior to the pixel based analysis method. The proposed method makes it possible to directly estimate the whole reservoir water area from, potentially, an individual cloud-free MODIS pixel, and is a promising way to improve the accuracy in the usability of MODIS images for the monitoring of reservoir surface water area variations.
... Water indices' techniques (Acharya et al., 2018) such as Normalized Difference Water Index (NDWI) (Xu, 2006), Automated Water Extraction Index (AWEI) (Feyisa et al., 2014), Modified NDWI (MNDWI) (Xu, 2006), Automated Water extraction Index (AWEI) (Feyisa et al., 2014), Normalized Difference Moisture Index (NDMI) (Gang and Dong-sheng, 2012;Gao, 1996;Hardisky et al., 1984), New Water Index (NWI) (Ding et al., 2018;Tian et al., 2017) and Water Ratio Index (WRI) (Shen and Li, 2010) are used widely. Classification methods including decision trees (DTs) (Acharya et al., 2016;Baker et al., 2006;Mueller et al., 2016;Tulbure and Broich, 2013;Tulbure et al., 2016), maximum likelihoods (MLs) (Frazier and Page, 2000;, statistical pattern recognition techniques (Acharya et al., 2018;Ji et al., 2015) are also used to acquire water bodies information. Whereas, recent advancement in automation various machine learning algorithms have been applied to extract water bodies from remote sensing images such as neural networks (NNs) (Rokni et al., 2015), artificial neural networks (ANN) (Skakun, 2012), Support vector machines (SVM) (Sun et al., 2014;Zhang et al., 2013), naive Bayes (NB), random forest (RF), gradient boosted machine (GBM), recursive partitioning and regression trees (RPART), and constraint energy minimizations (CEMs) (Acharya et al., 2019b;Ji et al., 2015;Wu et al., 2008). ...
Waterbody extraction in the high-elevated region of the Tibetan Plateau from remote sensing imagery is an efficient way to investigate and monitor water resources. The intrusion of shadows, snow/ice, and other impediments are still unconvincing in cryospheric regions. Landsat 8 OLI images were used to extract water bodies by applying various spectral indices such as; Normalized Difference Water Index (NDWI), Modified Normalized Difference Water Index (MNDWI), Normalized Difference Moisture Index (NDMI), Automated Water Extraction Index (AWEI_sh), New Water Index (NWI), Water Ratio Index (WRI), and LST based Water Extraction Index (LBWEI). The result showed that the LBWEI index has high accuracy in all weather conditions from 93.66% to 97.63% and improved consistency around 3% to 7% as compared to other techniques. The study suggests the LBWEI technique may be used in the cryospheric region for quick evaluation and provide the baseline information for researchers in algorithm developments.
... For the period of 1987 to 2020, all available Landsat (TM/ETM+/ OLI) surface reflectance products between June 26th and October 1st (dry season), where most of rain-fed agricultural areas were already harvested and shared similar spectral reflectance to barren areas, were acquired to extract irrigated areas. Since the available Landsat-5 could not cover the whole study area in 2003, 2005, and 2008.a), on the one hand, and Landsat-7 from June 2003 to December 2011 were reported unsuitable for extracting water bodies due to scan line corrector failure issue (Tulbure and Broich, 2013), on the other hand, we discarded the mentioned years from our analyses. Overall, 1,733 Landsat images, including 1,112 Landsat-5, 104 Landsat-7, and 517 Landsat-8 images were processed to analyze the annual changes in three classes of water, irrigated, and Non-Water/Irrigated (NWI) from 1987 to 2020 ( Fig. 2.c). ...
Irrigated agricultural expansion is one of the main reasons for water scarcity in the Lake Urmia basin. Although previous studies have analyzed the impact of cropland expansion on the Lake Urmia Shrinkage, there is a lack of comprehensive annual assessment of historical irrigation expansion in the Lake Urmia basin and its impact on water resources of this region. In this study, we developed an automatic and efficient workflow using Landsat and Gravity Recovery and Climate Experiment (GRACE) data, GRACE Follow-On (GRACE-FO) data, and a sample migration technique within the Google Earth Engine cloud computing platform to comprehensively investigate the impact of irrigated agricultural expansion on the shrinkage of Lake Urmia, as one of the most severe environmental crisis in the world. Additionally, using the global surface water data, we proposed a fully automatic procedure to obtain reference samples from water bodies. The Lake Urmia basin was first classified into the water, irrigated, and Non-Water/Irrigated classes using the random forest algorithm. The average overall accuracy of the produced annual land cover maps during 1987–2020 was 92.2%, representing the great potential of the developed method for land cover mapping. We found that the irrigated lands expanded by nearly 890 km² during the study period. Coincident with this change, although the area of water bodies in Lake Urmia partially recovered after 2015 (reached from 1,050 km² in 2015 to 3,370 km² in 2020), it is currently far beyond its original condition (i.e., ∼5,400 km², average record during 1987–2000). Moreover, the information of the Terrestrial Water Storage (TWS) from the GRACE and GRACE-FO data between 2003 and 2020 showed a dramatic decrease in TWS level (∼−11.5 cm). The findings of this research will assist the local stakeholders and authorities to better understanding the environmental costs of irrigation expansion in the Lake Urmia basin.
... In particular, they include spectral indices that allow strengthening indicators of the studied properties of water bodies [11,12]. To assess the possibilities of their use in the tasks of identifying pollution zones of estuaries of northern rivers, the following indices were considered in the work [13][14][15][16][17][18]: ...
The article discusses the tasks of creating a digital platform for integrated geoecolog-ical research based on space observations of the effects of natural and man-made sources of pollution of coastal marine areas. Data on the state of pollution of estuaries of the northern seas are given. Methods are proposed for identifying sections of surface water pollution from survey data using Landsat -8 space station. In this work, studies were conducted on the selection of spectral ranges of surveys in the tasks of isolating pollution of coastal surface waters. Examples are presented that illustrate the capabilities of satellite imagery in the tasks of identifying contaminated sites for estuarine areas of the marine areas of the Arctic Ocean. Recommendations are given on choosing the spectral ranges of surveys when conducting geoecological studies of marine waters.
... Hence, it is a negative indicator in the determination of SLSI. Similarly, waterbodies have a strong relationship with climatic changes having great impacts on anthropogenic water resources Pekel et al., 2016;Tao et al., 2015;Tulbure and Broich, 2013;Tulbure et al., 2014;Zou et al., 2017). Therefore, it is very much essential to protect waterbodies for environmental as well as anthropogenic purposes. ...
This is a comprehensive resource that integrates the application of innovative remote sensing techniques and geospatial tools in modeling Earth systems for environmental management beyond customary digitization and mapping practices. It identifies the most suitable approaches for a specific environmental problem, emphasizes the importance of physically based modeling, their uncertainty analysis, advantages, and disadvantages. The case studies on the Himalayas with a complex topography call for innovation in geospatial techniques to find solutions for various environmental problems.
Features:
Presents innovative geospatial methods in environmental modeling of Earth systems.
Includes case studies from South Asia and discusses different processes and outcomes using spatially explicit models.
Explains contemporary environmental problems through the analysis of various information layers.
Provides good practices for developing countries to help manage environmental issues using low-cost geospatial approaches.
Integrates geospatial modeling with policy and analysis its direct implication in decision making.
Using a systems’ approach analysis, Geospatial Modeling for Environmental Management: Case Studies from South Asia shall serve environmental managers, students, researchers, and policymakers.
... Several factors have come together to now allow for systematically mapping land cover at medium to high spatial resolution (10s of metres) across regional (Broich et al., 2011;Ghorbanian et al., 2020;Tulbure & Broich, 2013;Tulbure et al., 2016), national (Griffiths et al., 2019;Yang et al., 2020) and global extents (Buchhorn et al., 2020;Chen et al., 2017;GLanCE, 2021;Hansen et al., 2013;Pesaresi et al., 2016;Pickens et al., 2020;Potapov et al., 2021;Worldcover, 2021;Zhang et al., 2020Zhang et al., , 2021. Among the most important of these factors is the availability of freely accessible multi-temporal remote sensing datasets in a user-friendly format as 'Analysis Ready Data', ARD (Claverie et al., 2018;Dwyer et al., 2018;Frantz, 2019;Truckenbrodt et al., 2019;Wulder et al., 2016), such as the Committee on Earth Observation Satellites ARD for Land (CARD4L, https://ceos.org/ard/). ...
Unprecedented amounts of analysis‐ready Earth Observation (EO) data, combined with increasing computational power and new algorithms, offer novel opportunities for analysing ecosystem dynamics across large geographic extents, and to support conservation planning and action. Much research effort has gone into developing global EO‐based land‐cover and land‐use datasets, including tree cover, crop types, and surface water dynamics. Yet there are inherent trade‐offs between regional and global EO products pertaining to class legends, availability of training/validation data, and accuracy. Acknowledging and understanding these trade‐offs is paramount for both developing EO products and for answering science questions relevant for ecology or conservation studies based on these data. Here we provide context on the development of global EO‐based land‐cover and land‐use datasets, and outline advantages and disadvantages of both regional and global datasets. We argue that both types of EO‐derived land‐cover datasets can be preferable, with regional data providing the context‐specificity that is often required for policy making and implementation (e.g., land‐use and management, conservation planning, payment schemes for ecosystem services), making use of regional knowledge, particularly important when moving from land cover to actors. Ensuring that global and regional land‐cover and land‐use products derived based on EO data are compatible and nested, both in terms of class legends and accuracy assessment, should be a key consideration when developing such data. Open access high‐quality training and validation data derived as part of such efforts are of utmost importance. Likewise, global efforts to generate sets of essential variables for climate change, biodiversity, or eventually land use, which often require land‐cover maps as inputs, should consider regionalized, hierarchical approaches to not sacrifice regional context. Global change impacts manifest in regions, and so must the policy and planning responses to these challenges. EO data should embrace that regions matter, perhaps more than ever, in an age of global data availability and processing.
... Studies have generally quantified the changes in surface water extent (Alsdorf et al., 2007;Pekel et al., 2016), which is not sufficient to derive any action plans unless we have detailed spatial data on the size, location, trends of change, and their causes. The seasonal variation also accounts for changes in the extent of water bodies, viz., highest in winter and lowest in summer (Tulbure & Broich, 2013). Besides, the freely available spatial data sets of different periods (from 1970 to 1980s) had low resolution, which remains the major constraint to map small reservoirs/water bodies through automation. ...
Concerns have been raised about the threat of ecological imbalance due to the loss of water bodies in densely populated areas. The present study explored the changes in water bodies in terms of area, number, and size in northern districts of Tamil Nadu, India, between 1978 and 2018 using satellite data, geographic information system, spatial analysis, ground truth verification, and field validation. The analysis indicated that the water bodies’ area has reduced by 3027 ha and 4363 ha in the Kancheepuram and Tiruvallur Districts, respectively. Almost 179 water bodies have entirely disappeared, and 628 water bodies have been partly converted for other purposes. Of the disappeared water bodies, small, medium, and large water bodies account for 53, 93, and 33, respectively. The main reason for the changes in water bodies was the conversion to agriculture and buildings. Overall, the water bodies’ area and number have been reduced by 9% and 12%, respectively, while the population has grown by 37%. The water bodies lost due to anthropogenic activities demand the scientific inventory of water bodies and integrated water resources management at a state or national level with strict monitoring regulations to protect them.
... Like NDVI is highly sensitive to the brightness of soil and atmospheric effect, so we have considered some other indices, viz, MSAVI, EVI, ARVI, and NBR along with NDVI for more accurate classification of vegetation. Similarly, for identification of water features, water index like the NDWI (normalized difference water index), MNDWI (modified NDWI) [28,29,30,31,32] are considered. Built-up index enhances the spectral characteristics of built-up areas [33]. ...
Land use and land cover (LULC) provides a way to classify objects on the surface of Earth. This paper aims to identify the varying land cover classes by stacking of 6 spectral bands and 10 different generated indices from those bands together. We have considered the multispectral images of Landsat 7 for our research. It is seen that instead of using only basic spectral bands (blue, green, red, nir, swir1 and swir2) for classification, stacking relevant indices of multiple target classes like ndvi, evi, nbr, BU, etc. with basic bands generates more precise results. In this study, we have used automated clustering techniques for generating 5 different class labels for training the model. These labels are further used to develop a predictive model to classify LULC classes. The proposed classifier is compared with the SVM and KNN classifiers. The results show that this proposed strategy gives preferable outcomes over other techniques. After training the model over 50 epochs, an accuracy of 93.29% is achieved. Keywords: Land use, land cover, CNN, ISODATA, indices
... The Landsat 8 top-of-atmosphere (TOA) image was adopted in this study for consistency with other surface water mapping studies (Pekel et al., 2016;Tulbure and Broich, 2013). 93 images accessed from the GEE with path 124 and row 39 (see Fig. 1a) in the Worldwide Reference System-2 (WRS-2), acquired from April 2013 to May 2021, were selected. ...
Optical satellite-derived surface water monitoring is challenging because of the spatial gaps in images caused by clouds, cloud shadows, voids, etc. Here, an efficient method for filling gaps in time-series surface water images is proposed, based on the spatiotemporal characteristics of water. This method utilises the accurately classified historical ternary (gap, water, non-water) or binary (water, non-water) water image time-series and the clear part of the ternary gap water image. Pixels with values of 0 and 1 in the same period water occurrence image are first used to correct the gap water image. The spatial neighbourhood similarity is then calculated as a quality control band for mosaicking the accurately classified historical water images. The final result is generated by replacing the gap pixels with a mosaic image. The proposed method was implemented on the Google Earth Engine, and 93 Landsat 8 top-of-atmosphere (TOA) images were used to verify its validity. Quantitative evaluations were adequate, with a mean accuracy, recall, and precision of 0.98, 0.90, and 0.85, respectively. The proposed method could improve the utilisation of optical remote sensing data and would be applicable to the production of large-area homogeneous surface water time-series and water resource monitoring.
... Like NDVI is highly sensitive to the brightness of soil and atmospheric effect, so we have considered some other indices, viz, MSAVI, EVI, ARVI, and NBR along with NDVI for more accurate classification of vegetation. Similarly, for identification of water features, water index like the NDWI (normalized difference water index), MNDWI (modified NDWI) [28,29,30,31,32] are considered. Built-up index enhances the spectral characteristics of built-up areas [33]. ...
Land use and land cover (LULC) provides a way to classify objects on the surface of Earth. This paper aims to identify the varying land cover classes by stacking of 6 spectral bands and 10 different generated indices from those bands together. We have considered the multispectral images of Landsat 7 for our research. It is seen that instead of using only basic spectral bands (blue, green, red, nir, swir1 and swir2) for classification, stacking relevant indices of multiple target classes like ndvi, evi, nbr, BU, etc. with basic bands generates more precise results. In this study, we have used automated clustering techniques for generating 5 different class labels for training the model. These labels are further used to develop a predictive model to classify LULC classes. The proposed classifier is compared with the SVM and KNN classifiers. The results show that this proposed strategy gives preferable outcomes over other techniques. After training the model over 50 epochs, an accuracy of 93.29% is achieved.
... The Landsat mission, including the satellites and multi-spectral instruments from Landsat 4, 5, 7, and 8, represents the longest, continuous remote sensing monitoring dataset and is ideally suited for tracking multi-decadal lake dynamics (Wulder et al., 2016) as illustrated by its past and present use in numerous long-term water detection studies (Busker et al., 2019;Mueller et al., 2016;Pekel et al., 2016;Tang et al., 2016;Tulbure & Broich, 2013;Tulbure et al., 2016;Verpoorter et al., 2014;Weekley & Li, 2019;Yao et al., 2019). Despite possessing a modest 16-day to revisit period, overlap along scene edges will provide additional measurements for several water bodies, and the 30-m spatial resolution will enable observation of smaller water bodies with greater accuracy than possible with other sensors featuring better temporal resolution but decreased spatial resolution, like MODIS (Keys & Scott, 2018;Khandelwal et al., 2017;Moradi et al., 2014). ...
Identifying patterns and trends in long‐term lake dynamics is essential to establish effective water management procedures and boost our understanding of inland water's role in the global water cycle. This research leverages Google Earth Engine to estimate multi‐decadal water surface elevations for 52 lakes and reservoirs with varying physical properties. Water surface elevation was estimated using the entire Landsat 4, 5, 7, and 8 Landsat Top‐of‐Atmosphere Tier‐1 Collection‐1 archive from August 1982 through December 2017 via shoreline boundary statistics extracted from the National Elevation Dataset merged with lake bathymetry. Image contamination was identified and removed to provide elevation estimates for images with varying levels of image contamination. To improve accuracy, data filtering techniques were identified which retained over 70% of images with detectable water boundaries producing 26 lakes with sub‐meter root‐mean‐square‐error accuracy and 40 lakes with sub‐meter mean‐absolute‐error‐accuracy using a general overall parameter model. Additionally, lake‐specific locally optimized models were also determined with 45 of the 52 lakes producing sub‐meter root‐mean‐square‐error accuracies and 49 with sub‐meter mean‐absolute‐errors with individual lake accuracy as low as 0.191 m RMSE CI95%[0.129, 0.243]. In general, individual lake accuracy is highly correlated with the mean slope of the surrounding terrain with low‐slope shorelines having greater accuracy than high‐slope shorelines. Seasonal patterns in estimate accuracy were also identified. This research extends our ability to track lake dynamics over long time periods to lakes lacking traditional in‐situ monitoring, enables rapid assessment of lake dynamics across large areas, and balances a need for both high‐accuracy measurements and maximum temporal resolution.
... Recently, hydrological analyses of large areas have been trending toward more data-driven empirical approaches, because satellite imagery is the only way to assess water systematically over large spatial and temporal scales (Palazzoli & Ceola, 2020;Pekel et al., 2016;Perin et al., 2021;Tulbure & Broich, 2019;Wada et al., 2017;Walker et al., 2020). Surface-water-specific data sets derived from moderate resolution (30 m, Landsat) satellite imagery over 30-40 years are a relatively new development at the regional (Tulbure et al., 2016;Tulbure & Broich, 2013), national (Jones, 2015(Jones, , 2019, and global scales (Pekel et al., 2016;Pickens et al., 2020). The spatial and temporal scale of these surface water data sets, and the similar spatial and temporal scales of LULC (the Cropland Data Layer, CDL; "CropScape -NASS CDL Program"), population (LandScan; Rose et al., 2020), and climate data (Gridded Surface Meteorological data set; Abatzoglou, 2013), enable us to tackle the critical task of assessing the impact of climate and anthropogenic drivers on surface water. ...
Surface water is the most readily accessible water resource and provides an array of ecosystem services, but its availability and access are stressed by changes in climate, land cover, and population size. Understanding drivers of surface water dynamics in space and time is key to better managing our water resources. However, few studies estimating changes in surface water account for climate and anthropogenic drivers both independently and together. We used 19 years (2000–2018) of the newly developed Dynamic Surface Water Extent Landsat Science Product in concert with time series of precipitation, temperature, land cover, and population size to statistically model maximum seasonal percent surface water area as a function of climate and anthropogenic drivers in the southeastern United States. We fitted three statistical models (linear mixed effects, random forests, and mixed effects random forests) and three groups of explanatory variables (climate, anthropogenic, and their combination) to assess the accuracy of estimating percent surface water area at the watershed scale with different drivers. We found that anthropogenic drivers accounted for approximately 37% more of the variance in the percent surface water area than the climate variables. The combination of variables in the mixed effects random forest model produced the smallest mean percent errors (mean −0.17%) and the highest explained variance (R² 0.99). Our results indicate that anthropogenic drivers have greater influence when estimating percent surface water area than climate drivers, suggesting that water management practices and land‐use policies can be highly effective tools in controlling surface water variations in the Southeast.
... Based on the above strategies, previous research targeted different thematic LCCs, including changes in forest (Beamish et al., 2020;Fortin et al., 2020;Hansen et al., 2013;Richards et al., 2020), surface water (Berhane et al., 2020;Pekel et al., 2016;Tulbure and Broich, 2013), wetlands (Weise et al., 2020), croplands (Nguyen et al., 2020;Pérez-Hoyos et al., 2017;Serrano et al., 2019), grasslands , urban areas (Midekisa et al., 2017;Reba and Seto, 2020), and so on. So far, most efforts have concentrated on forest change monitoring at local, national (e.g., United States), regional (e.g., tropical forest area), and global scales using time series spectral bands and vegetation indices (VIs) derived from Landsat series satellites. ...
Comparing the performance of different satellite sensors in global land cover change (LCC) monitoring is necessary to assess their potential and limitations for more accurate and operational LCC estimations. This paper aims to examine and compare the performance in LCC monitoring using three satellite sensors: PROBA-V, Landsat 8 OLI, and Sentinel-2 MSI. We utilized a unique set of global reference data containing four years of records (2015-2018) at 29,263 land cover change/no-change 100 × 100-m sites. The LCC monitoring was conducted using the BFAST(s)-Random Forest (BRF) change detection framework involving 15 global timeseries vegetation indices and three BFAST models. Due to the different spectral characteristics and data availability of the sensors, we designed 30 comparison scenarios to extensively evaluate their performance. The overall results were: 1) for global general LCC monitoring, Landsat 8 OLI slightly outperformed Sentinel-2, and PROBA-V performed the worst. The performance among the three sensors differed consistently despite different data availability and spectral observation regions. Sentinel-2 was more competitive with Landsat 8 when the red-edge 1 band was included; 2) Landsat 8 was more accurate in forest, herbaceous vegetation, and water monitoring. Sentinel-2 performed particularly well in wetland monitoring. In addition, we further observed: 3) missing data in time series decreased the accuracy in all sensors, but had little influence on the relative performance across sensors; 4) combining sensors would not necessarily improve the accuracy because the complementary effects enhanced the accuracy only when there was a large amount of data missing for all sensors; 5) the BRF framework maintained the performance gap among sensors, but obtained a higher and more balanced accuracy overall when compared with using BFAST methods alone, by involving ensemble learning with an embedded sample-balancing strategy; 6) among the random forest variables, the 'magnitude' proved to be the most important contributor, and the NDVI had the most consistently good performance across sensors when compared against other vegetation indices. All sensors using BRF still had some errors in change detection, with a tendency to underestimate the global LCC. A potential reason for this is the complexity of the diverse change/no-change characteristics at the global extent and the fact that smaller, more subtle LCCs might not be well detected. These limitations could be addressed by taking advantage of ensemble learning approaches with a combination of multiple independent region/thematic-adapted LCC monitoring models and using the original Sentinel-2 (10 m) and Landsat 8 (30 m) in the future.
Background: Recent developments in optical satellite remote sensing have led to a new era in the detection of surface water with its changing dynamics. This study presents the creation of surface water inventory for a part of Pune district (an administrative area), in India using the Landsat 8 Operational Land Imager (OLI) and a multi spectral water indices method.
Methods: A total of 13 Landsat 8 OLI cloud free images were analyzed for surface water detection. Modified Normalized Difference Water Index (MNDWI) spectral index method was employed to enhance the water pixels in the image. Water and non-water areas in the map were discriminated using the threshold slicing method with a trial and error approach. The accuracy analysis based on kappa coefficient and percentage of the correctly classified pixels was presented by comparing MNDWI maps with corresponding Joint Research Centre (JRC) Global Surface Water Explorer (GSWE) images. The changes in the surface area of eight freshwater reservoirs within the study area (Bhama Askhed, Bhatghar, Chaskaman, Khadakwasala, Mulashi, Panshet, Shivrata, and Varasgaon) for the year 2016 were analyzed and compared to GSWE time series water databases for accuracy assessment. The annual water occurrence map with percentage water occurrence on a yearly basis was also prepared.
Results: The kappa coefficient agreement between MNDWI images and GSWE images is in the range of 0.56 to 0.96 with an average agreement of 0.82 indicating a strong level of agreement.
Conclusions: MNDWI is easy to implement and is a sufficiently accurate method to separate water bodies from satellite images. The accuracy of the result depends on the clarity of image and selection of an optimum threshold method. The resulting accuracy and performance of the proposed algorithm will improve with implementation of automatic threshold selection methods and comparative studies for other spectral indices methods.
Resumo Nos últimos anos, em meio à seca e a crise hídrica que afetou várias regiões do Brasil, em particular a região semiárida, os volumes dos reservatórios vêm sendo constantemente monitorados. Nesse contexto, o objetivo deste trabalho foi analisar, por meio de Sensoriamento Remoto, a dinâmica dos espelhos d'água dos reservatórios da Bahia, a fim de mostrar como a área dos espelhos d'água foram afetados pelas baixas precipitações, compreendendo os anos de 2012 a 2017. Para isso, foi utilizado a plataforma Google Earth Engine para analisar imagens do Landsat. Para a delimitação das águas, foi utilizada uma técnica de realce para converter as imagens RGB para HVS, criando uma imagem pancromática e facilitando o processo de identificação dos espelhos d'água. Desse modo, os resultados indicaram que a influência da variabilidade da precipitação e os impactos de outros fatores reduziram a quantidade de água superficial disponível de modo que dos 34 reservatórios estudados 16 tiveram redução de sua área ao final do período analisado. Essas informações são extremamente importantes para o planejamento e a gestão ambiental dos recursos hídricos, sob a perspectiva de fomentar políticas de abastecimento e, com isso, ampliar a capacidade de enfrentar problemas relacionados à segurança hídrica.
The adverse effects of flood disasters in urban areas have been increasing in severity and extent over the past years. The amounts of losses resulting from these events are also increasing exponentially, particularly in highly urbanised urban areas, where the effects of intensive land use and climate change are particularly extreme. All this despite our scientific knowledge, technical competence and computational capacity to develop highly sophisticated and accurate forecasting and simulation models being higher than ever, as is our capacity to map and analyse flood-related data.
In order to tackle this global issue, it is fundamental to keep on promoting and developing fundamental and applied research that allows the better targeting of interventions to improve resilience, reduce vulnerability and enhance recovery, as well as assisting decision-makers in delivering more effective flood risk-reduction policies. The present book aims to contribute to this goal by providing a space in which to share and discuss recent studies and state-of-the-art methodologies focused on the assessment and mitigation of flood risk in urban areas. It includes nine high-quality research articles authored by eminent scholars from India, Italy, Korea, Portugal, Romania, Singapore, Spain, Taiwan, Thailand and Vietnam, who had the tremendous generosity to join me in this project. The range of topics covered by these nine studies is extraordinarily vast, reflecting the complexity of the current challenges associated with the topic.
Studies on land use and land cover changes (LULCC) have been a great concern due to their contribution to the policies formulation and strategic plans in different areas and at different scales. The LULCC when intense and on a global scale can be catastrophic if not detected and monitored affecting the key aspects of the ecosystem's functions. For decades, technological developments and tools of geographic information systems (GIS), remote sensing (RS) and machine learning (ML) since data acquisition, processing and results in diffusion have been investigated to access landscape conditions and hence, different land use and land cover classification systems have been performed at different levels. Providing coherent guidelines, based on literature review, to examine, evaluate and spread such conditions could be a rich contribution. Therefore, hundreds of relevant studies available in different databases (Science Direct, Scopus, Google Scholar) demonstrating advances achieved in local, regional and global land cover classification products at different spatial, spectral and temporal resolutions over the past decades were selected and investigated. This article aims to show the main tools, data, approaches applied for analysis, assessment, mapping and monitoring of LULCC and to investigate some associated challenges and limitations that may influence the performance of future works, through a progressive perspective. Based on this study, despite the advances archived in recent decades, issues related to multi-source, mul-ti-temporal and multi-level analysis, robustness and quality, scalability need to be further studied as they constitute some of the main challenges for remote sensing.
This study presents an automated methodology to generate training data for surface water mapping from a single Sentinel-2 granule at 10 m (4 band, VIS/NIR) or 20 m (9 band, VIS/NIR/SWIR) resolution without the need for ancillary training data layers. The 20 m method incorporates an ensemble of three spectral indexes with optimal band thresholds, whereas the 10 m method achieves similar results using fewer bands and a single spectral index. A spectrally balanced and randomly generated set of training data based on the index values and optimal thresholds is used to fit machine learning classifiers. Statistical validation compares the 20 m ensemble-only method to the 20 m ensemble method with a random forest classifier. Results show the 20 m ensemble-only method had an overall accuracy of 89.5% (±1.7%), whereas the ensemble method combined with the random forest classifier performed better, with a ~4.8% higher overall accuracy: 20 m method (94.3% (±1.3%)) with optimal spectral index and SWIR thresholds of −0.03 and 800, respectively, and 10 m method (93.4% (±1.5%)) with optimal spectral index and NIR thresholds of −0.01 and 800, respectively. Comparison of other supervised classifiers trained automatically with the framework typically resulted in less than 1% accuracy improvement compared with the random forest, suggesting that training data quality is more important than classifier type. This straightforward framework enables accurate surface water classification across diverse geographies, making it ideal for development into a decision support tool for water resource managers.
Surface water plays a vital role in natural environment and human development. The research of water extraction method using remote sensing image is a hot topic, which has been widely developed in water index, classification, subpixel, and other aspects. Compared with other methods, a water-index based method has the advantages of fast speed and convenience. The characteristics of surface water, such as wide coverage and instability, make the water index stand out in monitoring large area of surface water. However, land surface in water environment is complex, and the main factors that reduce water extraction accuracy are also different, such as shadow in urban areas and water leakage in unshaded areas. The current index is bound to weaken the information in the water body when suppressing shadows, and vice versa. To address these issues, contrast difference water index (CDWI) and shadow difference water index (SDWI) are proposed in this paper by improving the modified normalized difference water index (MNDWI). CDWI is used to enhance water information, which is suitable for areas without building shadows. SDWI is used to eliminate the shadow of buildings, which is suitable for urban areas. Moreover, background difference water index (BDWI) was proposed by combining the advantages of CDWI and SDWI through a background regularizer B, which is used to extract surface water under complex background. The regularizer B represents the similarity between local background features of the image and the reference urban area, which is used to locally weight SDWI and CDWI, so that BDWI can automatically enhance the water body in the shadowless area and eliminate the shadow of buildings. The water extraction results of BDWI, MNDWI, the tasseled cap wetness index (TCW), the automatic water extraction index (AWEInsh, AWEIsh), and the water index 2015 (WI2015) were used for comparison. Other methods tend to perform well only in built-up areas or non-built-up areas, while the BDWI can extract surface water under various backgrounds with high accuracy and stability. The overall accuracy produced by the BDWI was 91.58–97.57%, CDWI was 84.85–97.09%, SDWI was 81.63–94.40%, MNDWI was 80.19–95.64%, TCW was 82.33–95.98%, AWEIsh was 87.50–96.37%, AWEInsh was 80.59–98.78%, and WI2015 was 78.24–98.38%. Combining water index with image local information is helpful to improve the accuracy of water extraction in large and complex environment. Finally, surface water in Jiangsu Province, China was extracted through BDWI and the changes in 1985, 2000, and 2015 were analyzed.
Abstract -
Surface water is the most readily accessible water resource and provides an array of ecosystem services, but is stressed by changes in climate, land cover, and population size. Understanding drivers of surface water dynamics in space and time is key to better managing our water resources. However, few studies estimating changes in surface water account for climate and anthropogenic drivers both independently and together. We used 19 years (2000-2018) of the newly developed Dynamic Surface Water Extent Landsat Science Product in concert with time series of precipitation, temperature, land cover, and population size to statistically model maximum seasonal percent surface water area as a function of climate and anthropogenic drivers in the Southeastern U.S. We fitted three statistical models (linear mixed effects, random forests, and mixed effects random forests) and three groups of explanatory variables (climate, anthropogenic, and their combination) to assess the accuracy of estimating percent surface water area at the watershed scale with different drivers. We found that anthropogenic drivers accounted for approximately 37% more of the variance in the percent surface water area than the climate variables. The combination of variables in the mixed effects random forest model produced the smallest mean percent errors (mean -0.17%) and the highest explained variance (R2 0.99). Our results indicate that anthropogenic drivers have greater influence when estimating percent surface water area than climate drivers, suggesting that water management practices and land use policies can be highly effective tools in controlling surface water variations in the Southeastern U.S.
We present a comprehensive critical review of well-established satellite remote sensing water indices and offer a novel, robust Augmented Normalized Difference Water Index (ANDWI). ANDWI employs an expanded set of spectral bands, RGB, NIR, and SWIR1-2, to maximize the contrast between water and non-water pixels. Further, we implement a dynamic thresholding method, the Otsu algorithm, to enhance ANDWI’s performance. Applied to a variety of environmental conditions, ANDWI with Otsu-thresholding offered the highest overall accuracy (accuracy=0.98, F1=0.98, and Kappa=0.96) compared to other indices (NDWI, MNDWI, AWEI, WI). We also propose a novel cloud filtering algorithm that substantially increases the number of usable images compared to the conventional cloud-free composites (124% increased observations in the studied area) and resolves inappropriate masking of water bodies and hot sands as clouds by conventional methods. Finally, we develop a Google Earth Engine App to readily delineate 16-day surface water bodies across the globe.
Rapid and accurate monitoring of irrigation dynamics in paddy fields (the start, end, duration and irrigation peak, etc.) at field scales is crucial to the fine management of agricultural water resources, especially in typical areas with water shortages. However, there is still a lack of sufficient research to depict irrigation dynamics in paddy fields at high temporal and spatial levels. To this end, this study fused Sentinel-2 and MODIS images to map the spatio-temporal dynamics of irrigation events in paddy fields. A popular spatiotemporal fusion algorithm (enhanced spatial and temporal adaptive reflectance fusion model, ESTARFM) was used to generate 25 high-spatial resolution (10 m) remote sensing images based on 9 Sentinel-2 images and 24 MODIS images. Random forest algorithm was used to extract the spatial distribution of irrigated paddy fields. Water body index and vegetation index were employed to identify the start, end and duration of irrigation in paddy fields. Penman-Monteith model was used to estimate water surplus and deficit of irrigation during the critical irrigation period with daily observation data from meteorological stations. This study was carried out in rice-growing areas in the middle and lower reaches of the Yellow River in China. The results indicated that the spatial distribution difference of irrigation events in paddy fields with the shortest 3-day interval could be monitored in collaboration with Sentinel-2 and MODIS images. The start, end, and duration of irrigation in paddy fields presented significant spatial differences at field scales. A large amount of water from groundwater and Yellow River was needed, because the total water shortage of irrigation in paddy fields in the study area was about 80.79% of the total water demand. In addition, paddy fields with higher water demand were more concentrated spatially, as were paddy fields with lower water demand. The feasibility of spatiotemporal fusion of multi-source remote sensing data makes it possible to continuously monitor irrigation dynamics in paddy fields on high spatial resolution scales, which is conducive to the construction of spatiotemporal database and big data platform of agricultural irrigation information. This would not only help in promoting the high-quality development of agricultural water resources management but also alleviating the contradiction of regional water resources.
The Prairie Pothole Region (PPR) contains numerous depressional wetlands known as potholes that provide habitats for waterfowl and other wetland-dependent species. Mapping these wetlands is essential for identifying viable waterfowl habitat and conservation planning scenarios, yet it is a challenging task due to the small size of the potholes, and the presence of emergent vegetation. This study develops an open-source process within the Google Earth Engine platform for mapping the spatial distribution of wetlands through the integration of Sentinel-1 C-band SAR (synthetic aperture radar) data with high-resolution (10-m) Sentinel-2 bands. We used two machine-learning algorithms (random forest (RF) and support vector machine (SVM)) to identify wetlands across the study area through supervised classification of the multisensor composite. We trained the algorithms with ground truth data provided through field studies and aerial photography. The accuracy was assessed by comparing the predicted and actual wetland and non-wetland classes using statistical coefficients (overall accuracy, Kappa, sensitivity, and specificity). For this purpose, we used four different out-of-sample test subsets, including the same year, next year, small vegetated, and small non-vegetated test sets to evaluate the methods on different spatial and temporal scales. The results were also compared to Landsat-derived JRC surface water products, and the Sentinel-2-derived normalized difference water index (NDWI). The wetlands derived from the RF model (overall accuracy 0.76 to 0.95) yielded favorable results, and outperformed the SVM, NDWI, and JRC products in all four testing subsets. To provide a further characterization of the potholes, the water bodies were stratified based on the presence of emergent vegetation using Sentinel-2-derived NDVI, and, after excluding permanent water bodies, using the JRC surface water product. The algorithm presented in the study is scalable and can be adopted for identifying wetlands in other regions of the world.
n recent years, in the midst of drought and water crisis that has affected several regions of Brazil, in particular the semi-arid region, surface water reserves have been constantly monitored. In this context, the objective of this study was to map and analyze, through Remote Sensing, the dynamics of the water mirrors of the main reservoirs in Ceará, in order to show how the area of the water mirrors of the reservoirs were affected by the precipitations below of the average during the last six years of drought, comprising the years 2012 to 2017. For this, the Google Earth Engine platform was used to analyze Landsat images, comprising the interannual period from 2012 to 2017. For the delimitation of the waters, an enhancement technique was used to convert the RGB images to HVS, creating a panchromatic image and facilitating the process of identifying the water mirrors. Thus, the results indicated that all reservoirs lost area, where some even dried up completely. The results also suggest that the reservoirs located in the hydrographic basins of the wetter climate showed less loss of area compared to those of the drier climate. Due to the high number of reservoirs, the use of satellite images and Remote Sensing techniques are essential to measure the effects of drought on dams. Such information is extremely important for the planning and environmental management of water resources, from the perspective of promoting supply policies and, with this, expanding the capacity to face problems related to water security.
en A system for the remote detection of water leaks in irrigation canals is presented which integrates visual, thermal, and multispectral imagers into a specially designed platform mounted on small unmanned aerial vehicles (UAV), along with an image processing and analysis procedure employing three different software packages and customized scripts. The end results are processed images of different colour bands and indexes which are shown to be effective at highlighting the leaks’ signatures. A custom payload platform weighing 1 kg was designed and built, consisting of high-resolution thermal, multispectral, and visual cameras, GPS units, a flight control camera, and a data acquisition system. A total of 27 flights in 4 irrigation districts were conducted to test and evaluate the payload platform and image processing system. The portable platform was successfully flown on three different UAV. A total of 46,387 images were captured and processed. The image processing and analysis system worked well and identified all known leaks, as well as leaks that the district's staff were unaware of. Thermal image indexes were effective at showing cold spots which were caused by canal leaks. The visual and the blue and green near-infrared indexes were also found to be effective at identifying leaks. The Normalized Difference Vegetation Index (NDVI) did help differentiate between vegetation and wet soils, but we did not obtain consistent results with the NDVI.
Résumé
fr Un système de détection à distance des fuites d'eau dans les canaux d'irrigation est présenté, qui intègre des caméras dans le domaine du visible, thermiques et multispectrales dans une plate-forme portable spécialement conçue pour être montée sur des drones aériens sans pilote pour collecter les images, avec une procédure de traitement et d'analyse d'images employant trois différents logiciels et codes personnalisés. Les images traitées obtenues des différentes bandes de couleurs et indices se sont révélées efficaces en mettant en évidence des signes des fuites d'eau sur les images. Une plateforme de charge utile personnalisée pesant 1 kg a été conçue et construite, comprenant des caméras thermiques, multispectrales et visuelles à haute résolution, des unités GPS, une caméra de contrôle de vol et un système d'acquisition de données. Un total de 27 vols ont été effectués dans 4 districts d'irrigation pour tester et évaluer la plate-forme de collecte et le système de traitement des images. La plateforme portable a été utilisée avec succès sur trois drones différents. Au total, 46,387 images ont été capturées et traitées. Le système de traitement et d'analyse des images a bien fonctionné et a identifié toutes les fuites d'eau connues, ainsi que les fuites inconnues dont le personnel du district n'était pas au courant. Les indices des images thermiques étaient les plus efficaces pour déterminer les « points froids » sur les images causés par des fuites d'eau de canal. Les indices visuels et les images infrarouges bleu et vertes se sont également avérés efficaces pour identifier les fuites. Le NDVI a aidé à différencier la végétation des sols humides dans certains cas, mais nous n'avons pas obtenu de résultats cohérents avec le NDVI.
This chapter presents an overview of the main time series analysis methods for environment monitoring with earth observation, from classical methods to the deep learning (DL) methods. It summarizes main differences between bi-temporal change detection, annual time series and dense time series analyses, and also presents the three main types of annual time series methods for environment monitoring. The chapter focuses on dense time series methods using all available data, first presenting the main data preprocessing requirements, and provides an overview of the four main types of change detection methods based on dense time series analysis. These include: map classification, trajectory classification, statistical boundary and ensemble approaches. The chapter discusses three kinds of network architectures suited for the analysis of satellite image time series (SITS): recurrent neural networks, convolutional neural networks and hybrid models combining both. It proposes a prospective reflection upon possible convergence at crossroads between SITS analysis, video processing, computer vision and DL.
Water monitoring is an important part of water resource protection. The extraction of water body from multispectral remote-sensing images has been proven to be an efficient and fast way for water monitoring. This paper presents a water body extraction algorithm from multispectral remote-sensing image based on region similarity and boundary information by combining adaptive band selection and over-segmentation. First of all, three bands are adaptively chosen by similarity-based band selection algorithm. Then, the image domain is partitioned into a series of homogeneous sub-regions by over-segmentation incorporating spectral and spatial information. On the sub-regions, the regional similarity is defined with respect to the similarities of texture and spectral features which are extracted using structure analysis method. After that, boundary information is extraction by Canny algorithm, then the water body is extracted by using the Fractal Net Evolution Approach (FNEA) which combines regional similarity and boundary information. The proposed algorithm is used to extract six water bodies with different complex texture backgrounds from multispectral sensors. According to the accuracy evaluation of water body extraction results, the overall accuracy (OA) is higher than 97.9100% and all Kappa coefficients (K) are up to 0.9436. We calculated the relative error (RE) of the area between the reference water body and the water body extracted by the proposed algorithm, the minimum and maximum relative error range is between [0.6180%, 7.7050%]. The experiments show that the proposed algorithm is feasible and effective.
Before being used in scientific investigations and policy decisions, thematic maps constructed from remotely sensed data should be subjected to a statistically rigorous accuracy assessment. The three basic components of an accuracy assessment are: 1) the sampling design used to select the reference sample; 2) the response design used to obtain the reference land-cover classification for each sampling unit; and 3) the estimation and analysis procedures. We discuss options available for each of these components. A statistically rigorous assessment requires both a probability sampling design and statistically consistent estimators of accuracy parameters, along with a response design determined in accordance with features of the mapping and classification process such as the land-cover classification scheme, minimum mapping unit, and spatial scale of the mapping.
Relatively few large-scale inventories of the world's wetlands exist because of the difficulties of spatial scale, associated cost and multiple objectives, often temporally confounded, that drive classification. The extent of wetlands across a large part of Australia (New South Wales, 80.6 million ha) was determined using satellite image analyses. These data allowed analyses of the distribution of wetlands, their conservation status and potential threats at different spatial scales; that is, State, coastal and inland, and catchment. Approximately 5.6% of New South Wales is wetland (4.5 million ha), mostly (96%) in inland river catchments. Broad classification allowed identification of the extent of wetland types: (i) floodplains (89%); (ii) freshwater lakes (6.6%); (iii) saline lakes (< 1%); (iv) estuarine wetlands (2.5%); and (v) coastal lagoons and lakes (1.5%). Conservation reserves protect only 3% of wetland area. The analyses identified the north-west as the key area for wetland conservation as most other catchments have lower wetland extent and more potential threatening processes. The first stage of a large-scale inventory is to determine the extent and location of wetlands, with immediate benefits for strategic conservation and management. Other objectives (e.g. classification, biotic composition, hydrology and threats) seldom have sufficient data available for large-scale inventories but can be completed later with resources.
The Indonesian islands of Sumatera and Kalimantan (the Indonesian part of the island of Borneo) are a center of significant and rapid forest cover loss in the humid tropics with implications for carbon dynamics, biodiversity conservation, and local livelihoods. The aim of our research was to analyze and interpret annual trends of forest cover loss for different sub-regions of the study area. We mapped forest cover loss for 2000–2008 using multi-resolution remote sensing data from the Landsat enhanced thematic mapper plus (ETM+) and moderate resolution imaging spectroradiometer (MODIS) sensors and analyzed annual trends per island, province, and official land allocation zone. The total forest cover loss for Sumatera and Kalimantan 2000–2008 was 5.39 Mha, which represents 5.3% of the land area and 9.2% of the year 2000 forest cover of these two islands. At least 6.5% of all mapped forest cover loss occurred in land allocation zones prohibiting clearing. An additional 13.6% of forest cover loss occurred where clearing is legally restricted. The overall trend of forest cover loss increased until 2006 and decreased thereafter. The trends for Sumatera and Kalimantan were distinctly different, driven primarily by the trends of Riau and Central Kalimantan provinces, respectively. This analysis shows that annual mapping of forest cover change yields a clearer picture than a one-time overall national estimate. Monitoring forest dynamics is important for national policy makers, especially given the commitment of Indonesia to reducing greenhouse gas emissions as part of the reducing emissions from deforestation and forest degradation in developing countries initiative (REDD+). The improved spatio-temporal detail of forest change monitoring products will make it possible to target policies and projects in meeting this commitment. Accurate, annual forest cover loss maps will be integral to many REDD+ objectives, including policy formulation, definition of baselines, detection of displacement, and the evaluation of the permanence of emission reduction.
Many wetlands (estimated to be about 70%) have been lost in the coastal plain region of southwestern Australia since British settlement (in 1829), primarily as a result of infilling or drainage to create land for agricultural use or urban development. While further loss is almost universally acknowledged as undesirable, wetland degradation continues with little overt public recognition of the causes or consequences. Obvious and direct causes include nutrient enrichment, salinization, pollution with pesticides and heavy metals, the invasion of exotic flora and flora, loss of fringing vegetation and altered hydrological regimes occurring as a result of urbanization and agricultural practices. Underlying causes include a lack of understanding of wetland hydrology and ecology on behalf of both planning agencies and the private sector, and poor coordination of the many different agencies responsible for wetland management. Public and political awareness of wetland values continues to increase, but sectoral organization and responsibilities for wetland management lag behind. Sufficient scientific information now exists for improved management, protection and restoration of wetlands in southwestern Australia. However, this improvement cannot occur without the necessary political will and corresponding sectoral responses needed to implement coordinated wetland management policies and actions.
Studies of wetlands in Australia, as in other countries, have taken a wide variety of approaches to defining, surveying and classifying these environments. Past and current approaches in Australia are reviewed for each of the States and Territories which provide the context for much of the natural resource investigation in the country. While there are obvious advantages of national, and perhaps international, agreement on definition and types of wetlands, a variety of approaches to inventory and classification will always be necessary for particular purposes. More fundamental than general agreement on approaches is the need for wetland scientists and managers to maximise the accuracy of survey information, to test the assumptions involved in the use of classifications, and to ensure that the classifications they use are the most appropriate for their purposes. The issue of a global wetland classification scheme is discussed on the basis of a representative range of views by Australian wetland workers.
Management of wetlands influenced by the Gnangara groundwater mound, Western Australia, is becoming increasingly challenging
due to an ongoing decline in the regional groundwater table. A number of these groundwater-dependent wetlands have acidified
(with adverse effects on the extant macroinvertebrate fauna) due to the oxidation of pyritic sediments. One management option
in such cases is artificial augmentation of surface water in order to maintain or reinstate anaerobia in the sediments. This
paper documents cycles of macroinvertebrate decline and recovery over 12 years of monitoring in three Gnangara mound wetlands
affected by drought-induced acidification, one of which is being artificially augmented. Acidification did not result in a
reduction of the total number of macroinvertebrate families present, however, there were clearly identifiable groups of acid-sensitive
taxa (amphipods and isopods, ostracods, chydorid and daphnid cladocerans, mayflies, oligochaetes, clams and snails) and acid-tolerant
taxa (sandfly larvae, macrothricid cladocerans and water boatmen). In the artificially augmented wetland, the effects of acidification
were reversed: acid-sensitive taxa reappeared and acid-tolerant taxa decreased in numbers. Moreover, there were a number of
taxa that appeared for the first time since augmentation, and summer family richness increased markedly. This study has shown
that artificial augmentation of wetland water levels can be a successful recovery strategy for recently acidified systems,
but this will depend on a number of factors, and the ‘recovered’ state will be at least slightly different from the original
state.
A geomorphic classification of inland wetlands on criteria other than vegetation is proposed, based on their host landform and degree of wetness. Thus, the classification addresses the underlying structure of most inland wetlands, i.e. their landform setting and their various types of hydroperiod. Landforms host to wetlands include: basins, channels, flats, slopes and hills/highlands. Degrees of wetness include: permanent, seasonal or intermittent inundation, and seasonal waterlogging. From combining the landform type with hydroperiod, thirteen primary types of common wetlands are recognized: 1. permanently inundated basin = lake; 2. seasonally inundated basin = sumpland; 3. intermittently inundated basin = playa; 4. seasonally waterlogged basin = dampland; 5. permanently inundated channel = river; 6. seasonally inundated channel = creek; 7. intermittently inundated channel = wadi; 8. seasonally waterlogged channel = trough; 9. seasonally inundated flat = floodplain; 10. intermittently inundated flat = barlkarra; 11. seasonally waterlogged flat = palusplain; 12. seasonally waterlogged slope = paluslope; and 13. seasonally waterlogged highlands = palusmont. Water, landform and vegetation descriptors can augment the nomenclature of the primary units: e.g. salinity of water; size and shape of landform; and organisation, structure and floristics of vegetation. The classification can be used in many settings, regardless of climate and vegetation types. Using the approach adopted in this classification, in principle, more landform types and degrees of wetness, if necessary, can be added to the system to define additional wetland types.
To conserve and manage wetland resources, it is important to inventoryand monitor wetlands and their adjacent uplands. Satellite remote sensing hasseveral advantages for monitoring wetland resources, especially for largegeographic areas. This review summarizes the literature on satellite remotesensing of wetlands, including what classification techniques were mostsuccessful in identifying wetlands and separating them from other land covertypes. All types of wetlands have been studied with satellite remote sensing.Landsat MSS, Landsat TM, and SPOT are the major satellite systems that have beenused to study wetlands; other systems are NOAA AVHRR, IRS-1B LISS-II and radarsystems, including JERS-1, ERS-1 and RADARSAT. Early work with satellite imageryused visual interpretation for classification. The most commonly used computerclassification method to map wetlands is unsupervised classification orclustering. Maximum likelihood is the most common supervised classificationmethod. Wetland classification is difficult because of spectral confusion withother landcover classes and among different types of wetlands. However,multi-temporal data usually improves the classification of wetlands, as doesancillary data such as soil data, elevation or topography data. Classifiedsatellite imagery and maps derived from aerial photography have been comparedwith the conclusion that they offer different but complimentary information.Change detection studies have taken advantage of the repeat coverage andarchival data available with satellite remote sensing. Detailed wetland maps canbe updated using satellite imagery. Given the spatial resolution of satelliteremote sensing systems, fuzzy classification, subpixel classification, spectralmixture analysis, and mixtures estimation may provide more detailed informationon wetlands. A layered, hybrid or rule-based approach may give better resultsthan more traditional methods. The combination of radar and optical data providethe most promise for improving wetland classification.
Accurate, efficient, and repeatable mapping of changes in wetlands and riparian areas (referred to collectively as wetlands)
is critical for monitoring human, climatic, and other effects on these important systems. We used Landsat-based satellite
imagery from 1988 and 2001 to map changes in wetland ecosystems in the Gallatin Valley of southwest Montana. Stochastic gradient
boosting (SGB) was used to classify the 2001 image, and change vector analysis (CVA) was used to identify locations where
wetland areas might have changed between 1988 and 2001. These potentially changed locations again were classified for the
1988 Landsat image using SGB. Areas of change constituted 3.4% of the study area, thus only this small percentage of the image
was reclassified for the 1988 image. Overall change detection accuracy was 76%, although changes along the periphery of wetland
boundaries and in areas of smaller upland inclusions were not distinguished as well as other changes. Overall accuracies of
the SGB wetland classification maps were 81% for 1988 and 86% for 2001. CVA significantly reduced the number of pixels involved
in the historical image classification compared to conducting independent classifications, thus reducing the potential for
compounding classification errors in unchanged areas.
The location and distribution of wetlands and riparian zones influence the ecological functions present on a landscape. Accurate
and easily reproducible land-cover maps enable monitoring of land-management decisions and ultimately a greater understanding
of landscape ecology. Multi-season Landsat ETM+ imagery from 2001 combined with ancillary topographic and soils data were
used to map wetland and riparian systems in the Gallatin Valley of Southwest Montana, USA. Classification Tree Analysis (CTA)
and Stochastic Gradient Boosting (SGB) decision-tree-based classification algorithms were used to distinguish wetlands and
riparian areas from the rest of the landscape. CTA creates a single classification tree using a one-step-look-ahead procedure
to reduce variance. SGB uses classification errors to refine tree development and incorporates multiple tree results into
a single best classification. The SGB classification (86.0% overall accuracy) was more effective than CTA (73.1% overall accuracy)
at detecting a variety of wetlands and riparian zones present on this landscape.
There has been an increasing interest in characterizing and mapping isolated depressional wetlands due to a 2001 U.S. Supreme
Court decision that effectively removed their protected status. Our objective was to determine the utility of satellite remote
sensing to accurately detect isolated wetlands. Image segmentation and object-oriented analysis were applied to Landsat-7
imagery from January and October 2000 to map isolated wetlands in the St. Johns River Water Management District of Alachua
County, Florida. Accuracy for individual isolated wetlands was determined based on the intersection of reference and remotely
sensed polygons. The January data yielded producer and user accuracies of 88% and 89%, respectively, for isolated wetlands
larger than 0.5 acres (0.20 ha). Producer and user accuracies increased to 97% and 95%, respectively, for isolated wetlands
larger than 2 acres (0.81 ha). Recently, the Federal Geographic Data Committee recommended that all U.S. wetlands 0.5 acres
(0.20 ha) or larger should be mapped using 1-m aerial photography with an accuracy of 98%. That accuracy was nearly achieved
in this study using a spatial resolution that is 900 times coarser. Satellite remote sensing provides an accurate, relatively
inexpensive, and timely means for classifying isolated depressional wetlands on a regional or national basis.
Since January 2008, the U.S. Department of Interior / U.S. Geological Survey have been providing free terrain-corrected (Level 1T) Landsat Enhanced Thematic Mapper Plus (ETM+) data via the Internet, currently for acquisitions with less than 40% cloud cover. With this rich dataset, temporally composited, mosaics of the conterminous United States (CONUS) were generated on a monthly, seasonal, and annual basis using 6521 ETM+ acquisitions from December 2007 to November 2008. The composited mosaics are designed to provide consistent Landsat data that can be used to derive land cover and geo-physical and bio-physical products for detailed regional assessments of land-cover dynamics and to study Earth system functioning. The data layers in the composited mosaics are defined at 30 m and include top of atmosphere (TOA) reflectance, TOA brightness temperature, TOA normalized difference vegetation index (NDVI), the date each composited pixel was acquired on, per-band radiometric saturation status, cloud mask values, and the number of acquisitions considered in the compositing period. Reduced spatial resolution browse imagery, and top of atmosphere 30 m reflectance time series extracted from the monthly composites, capture the expected land surface phenological change, and illustrate the potential of the composited mosaic data for terrestrial monitoring at high spatial resolution. The composited mosaics are available in 501 tiles of 5000 × 5000 30 m pixels in the Albers equal area projection and are downloadable at http://landsat.usgs.gov/WELD.php. The research described in this paper demonstrates the potential of Landsat data processing to provide a consistent, long-term, large-area, data record.
The southwestern corner of the Australian continent has been identified as a global "biodiversity hotspot", defined as an area where "exceptional concentrations of endemic species are undergoing exceptional loss of habitat". In this paper we reconsider the reasons for this conservation priority. We briefly review significant characteristics of the flora and fauna, and the way threatening processes are escalating ecosystem stress to these conservation values. Our specific aim is to examine the ecological consequences of hydrological change, including emergent issues such as climate change, and focus on the coastal plains in higher rainfall zones where the majority of the Western Australian population resides. Here we argue that human-driven and/or climatically driven hydrological change deserve greater attention, since they: i) directly escalate the risk of extinction for some components of the biota, or ii) are underlying and/or contributing factors in the manifestation of other threats to the biota, and may complicate or exacerbate some of those threats (such as fire, Phytophthora and the spread of weeds). This paper briefly outlines the challenges to the region's biodiversity posed by hydrological change. We suggest a societal adoption of approaches based on water literacy will be necessary to avoid irreversible changes associated with a continued reliance on water resource developments and other energy/water intensive industrial activities.
The Wetlands and Water synthesis was designed for the Ramsar Convention to meet the need for information about the consequences of ecosystem change for human well-being and sought to strengthen the link between scientific knowledge and decision-making for the conservation and wise use of wetlands.
The aim of this project was to determine the accuracy of using simple digital image processing techniques to map riverine water bodies with Landsat 5 TM data. This paper quantifies the classification accuracy of single band density slicing of Landsat 5 TM data to delineate water bodies on riverine floodplains. The results of these analyses are then compared to a 6-band maximum likelihood classification over the same area. The water boundaries delineated by each of these digital classification procedures were compared to water boundaries delineated from colour aerial photography acquired on the same day as the TM data. These comparisons show that Landsat TM data can be used to map water bodies accurately. Density slicing of the single mid-infrared band 5 proved as successful as multispectral classification achieving an overall accuracy of 96.9%, a producer's accuracy for water bodies of 81.7% and a user's accuracy for water bodies of 64.5%.
Landsat, first placed in orbit in 1972, established the U.S. as the world leader in land remote sensing. The Landsat system has contributed significantly to the understanding of the Earth’s environment, spawned revolutionary uses of space-based data by the commercial value-added
industry, and encouraged a new generation of commercial satellites that provide regional, high-resolution spatial images. This PE&RS Special Issue provides an update to the 1997 25th Landsat anniversary issue, particularly focused on the contribution of Landsat-7 to the 34+
year history of the Landsat mission. In this overview paper, we place the Landsat-7 system in context and show how mission operations have changed over time, increasingly exploiting the global monitoring capabilities of the Landsat observatory. Although considerable progress was made during
the Landsat-7 era, there is much yet to learn about the historical record of Landsat global coverage: a truly valuable national treasure. The time to do so is now, as the memories of the early days of this historic program are fading as we speak.
Surface fresh water is essential for life, yet we have surprisingly poor
knowledge of the spatial and temporal dynamics of surface freshwater
discharge and changes in storage globally. For example, we are unable to
answer such basic questions as "What is the spatial and temporal
variability of water stored on and near the surface of all continents?"
Furthermore, key societal issues, such as the susceptibility of life to
flood hazards, cannot be answered with the current global, in situ
networks designed to observe river discharge at points but not flood
events. The measurements required to answer these hydrologic questions
are surface water area, the elevation of the water surface (h), its
slope (∂h/∂x), and temporal change (∂h/∂t). Advances
in remote sensing hydrology, particularly over the past 10 years and
even more recently, have demonstrated that these hydraulic variables can
be measured reliably from orbiting platforms. Measurements of inundated
area have been used to varying degrees of accuracy as proxies for
discharge but are successful only when in situ data are available for
calibration; they fail to indicate the dynamic topography of water
surfaces. Radar altimeters have a rich, multidecadal history of
successfully measuring elevations of the ocean surface and are now also
accepted as capable tools for measuring h along orbital profiles
crossing freshwater bodies. However, altimeters are profiling tools,
which, because of their orbital spacings, miss too many freshwater
bodies to be useful hydrologically. High spatial resolution images of
∂h/∂t have been observed with interferometric synthetic
aperture radar, but the method requires emergent vegetation to scatter
radar pulses back to the receiving antenna. Essentially, existing
spaceborne methods have been used to measure components of surface water
hydraulics, but none of the technologies can singularly supply the water
volume and hydraulic measurements that are needed to accurately model
the water cycle and to guide water management practices. Instead, a
combined imaging and elevation-measuring approach is ideal as
demonstrated by the Shuttle Radar Topography Mission (SRTM), which
collected images of h at a high spatial resolution (˜90 m) thus
permitting the calculation of ∂h/∂x. We suggest that a future
satellite concept, the Water and Terrestrial Elevation Recovery mission,
will improve upon the SRTM design to permit multitemporal mappings of h
across the world's wetlands, floodplains, lakes, reservoirs, and rivers.
Landsat data constitute the longest record of global-scale medium spatial resolution earth observation data. As a result, the current methods for large area monitoring of land cover change using medium spatial resolution imagery (10–50 m) typically employ Landsat data. Most large area products quantify forest cover change. Forests are a comparatively easy cover type to map as well as a current focus of environmental monitoring concerning the global carbon cycle and biodiversity loss. Among existing change products, supervised or knowledge-based characterization methods predominate. Radiometric correction methods vary significantly, largely as a function of geographic/algorithmic scale. For instance, products created by mosaicking per scene characterizations do not require radiometric normalization. On the other hand, methods that employ a single index or classification model over an entire study area do require radiometric normalization. Temporal updating of cover change varies between existing products as a function of regional acquisition frequency, cloud cover and seasonality. With the Landsat archive opened for free access to terrain-corrected data, future product generation will be more data intensive. Per scene, interactive analyses will no longer be viable. Coupling free and open access to large data volumes with improved processing power will result in automated image pre-processing and land cover characterization methods. Such methods will need to leverage high-performance computing capabilities in advancing the land cover monitoring discipline. Robust validation efforts will be required to quantify product accuracies in determining the optimal change characterization methodologies.
The growing availability of multi-temporal satellite data has increased opportunities for monitoring large rivers from space. A variety of passive and active sensors operating in the visible and microwave range are currently operating, or planned, which can estimate inundation area and delineate flood boundaries. Radar altimeters show great promise for directly measuring stage variation in large rivers. It also appears to be possible to obtain estimates of river discharge from space, using ground measurements and satellite data to construct empirical curves that relate water surface area to discharge. Extrapolation of these curves to ungauged sites may be possible for the special case of braided rivers. Where clouds, trees and floating vegetation do not obscure the water surface, high-resolution visible/infrared sensors provide good delineation of inundated areas. Synthetic aperture radar (SAR) sensors can penetrate clouds and can also detect standing water through emergent aquatic plants and forest canopies. However, multiple frequencies and polarizations are required for optimal discrimination of various inundated vegetation cover types. Existing single- polarization, fixed-frequency SARs are not suÅcient for mapping inundation area in all riverine environments. In the absence of a space-borne multi-parameter SAR, a synergistic approach using single-frequency, fixed-polarization SAR and visible/infrared data will provide the best results over densely vegetated river floodplains. #1997 John Wiley & Sons, Ltd.
This assessment is based on reviews of the extent of wetland inventory in seven regions of the world. A key conclusion is that little is still known about the extent and condition of the global wetland resource. It was not possible to make reliable overall estimates of the size of the wetland resource globally or regionally. Previous estimates range from 5.3 to 9.7 million km(2) but present analyses now suggest a tentative minimum of 12.8 km(2). Recommendations focus on the need for national inventory programmes and the inclusion of basic information on the location and extent of each wetland and its major ecological features as a forerunner to collecting further management-oriented information. Thus, the following core data should be collected: area and boundary, location, geomorphic setting, general description, soil characteristics, water regime, water quality, and biotic characteristics. Further, the development of standardized methods for data collection, collation and storage are called for. These should address the use of remotely sensed data and storage of information in electronic formats, including Geographic Information Systems and recording key information in a meta-database. Habitats of priority for future inventory are seagrasses, coral reefs, salt marshes and coastal flats, mangroves, arid-zone wetlands, peatlands, rivers and streams, and artificial wetlands.
This study aimed to develop simple remote-sensing techniques suitable for mapping and monitoring
wetlands, using Landsat TM imagery of inland wetland sites in Victoria and New South Wales. A range
of classification methods was examined in attempts to map the location and extent of wetlands and their
vegetation types. Multi-temporal imagery (winter/spring and summer) was used to display seasonal
variability in water regime and vegetation status. Simple density slicing of the mid-infrared band (TM5)
from imagery taken during wet conditions was useful for mapping the location and extent of inundated
areas. None of the classification methods tested reproduced field maps of dominant vegetation species;
however, density slicing of multi-temporal imagery produced classes based on seasonal variation in
water regime and vegetation status that are useful for reconnaissance mapping and for examining
variability in previously mapped units. Satellite imagery is unlikely to replace aerial photography for
detailed mapping of wetland vegetation types, particularly where ecological gradients are steep, as in
many riverine systems. However, it has much to offer in monitoring changes in water regime and in
reconnaissance mapping at regional scales.
On 31 May 2003, the Landsat Enhanced Thematic Plus (ETM+) Scan Line Corrector (SLC) failed, causing the scanning pattern to exhibit wedge-shaped scan-to-scan gaps. We developed a method that uses coincident spectral data to fill the image gaps. This method uses a multi-scale segment model, derived from a previous Landsat SLC-on image (image acquired prior to the SLC failure), to guide the spectral interpolation across the gaps in SLC-off images (images acquired after the SLC failure). This paper describes the process used to generate the segment model, provides details of the gap-fill algorithm used in deriving the segment-based gap-fill product, and presents the results of the gap-fill process applied to grassland, cropland, and forest landscapes. Our results indicate this product will be useful for a wide variety of applications, including regional-scale studies, general land cover mapping (e.g. forest, urban, and grass), crop-specific mapping and monitoring, and visual assessments. Applications that need to be cautious when using pixels in the gap areas include any applications that require per-pixel accuracy, such as urban characterization or impervious surface mapping, applications that use texture to characterize landscape features, and applications that require accurate measurements of small or narrow landscape features such as roads, farmsteads, and riparian areas.
Summary1. Changes in land management (land use and land cover) and water management (including extraction of ground water and diversion of surface waters for irrigation) driven by increases in agricultural production and urban expansion (and fundamentally by population growth) have created multiple stressors on global freshwater ecosystems that we can no longer ignore.2. The development and testing of conceptual ecological models that examine the impact of stressors on aquatic ecosystems, and recognise that responses may be nonlinear, is now essential for identifying critical processes and predicting changes, particularly the possibility of catastrophic regime shifts or ‘ecological surprises’.3. Models depicting gradual ecological change and three types of regime shift (simple thresholds, hysteresis and irreversible changes) were examined in the context of shallow inland aquatic ecosystems (wetlands, shallow lakes and temporary river pools) in southwestern Australia subject to multiple anthropogenic impacts (hydrological change, eutrophication, salinisation and acidification).4. Changes in hydrological processes, particularly the balance between groundwater-dominated versus surface water-dominated inputs and a change from seasonal to permanent water regimes appeared to be the major drivers influencing ecological regime change and the impacts of eutrophication and acidification (in urban systems) and salinisation and acidification (in agricultural systems).5. In the absence of hydrological change, urban wetlands undergoing eutrophication and agricultural wetlands experiencing salinisation appeared to fit threshold models. Models encompassing alternative regimes and hysteresis appeared to be applicable where a change from a seasonal to permanent hydrological regime had occurred.6. Irreversible ecological change has potentially occurred in agricultural landscapes because the external economic driver, agricultural productivity, persists independently of the impact on aquatic ecosystems.7. Thematic implications: multiple stressors can create multiple thresholds that may act in a hierarchical fashion in shallow, lentic systems. The resulting regime shifts may follow different models and trajectories of recovery. Challenges for ecosystem managers and researchers include determining how close a system may be to critical thresholds and which processes are essential to maintaining or restoring the system. This requires an understanding of both external drivers and internal ecosystem dynamics, and the interactions between them, at appropriate spatial and temporal scales.