Operational Land Imager Imagery Research Articles

National-scale nighttime high-temperature anomalies from Landsat-8 OLI images

Identifying and characterizing the distribution, abundance, and structural composition of national high-temperature anomalies (HTAs) are essential for mitigating natural disasters, regulating industrial emissions, and assessing societal impacts. The Landsat-8 Operational Land Imager (OLI) provides a unique capability for monitoring nighttime HTAs (N-HTAs) at high spatial resolutions without contamination from reflected solar radiation. However, the extraction of N-HTAs from nighttime OLI imagery poses significant challenges due to the interference of bright up-light, especially for smaller and cooler hotspots, and the absence of spectral information on backgrounds complicates the classification of N-HTAs at night. Here, we explored the background noise characteristics of Bands 2–7 in nighttime OLI imagery and developed an object-oriented approach for the enhanced detection of small and cool N-HTAs. Furthermore, we introduced a deep learning model that leverages high-resolution aerial imagery and land cover data to categorize N-HTAs into industrial sources, living sources, vegetation fires, and others. Applying these methodologies to nighttime OLI imagery from 2013 to 2020 allowed us to map and classify N-HTAs across the contiguous United States (CONUS), facilitating the creation of a detailed spatiotemporal inventory of industrial N-HTAs. Our analysis revealed that N-HTAs from living sources dominate the total N-HTAs, with a pronounced holiday effect, and gas flaring activities in oil and gas development areas account for >85 % of industrial N-HTAs. The outcomes of this national-scale investigation underscore the significant potential and practical utility of nighttime OLI imagery for N-HTA monitoring and advocate for expanding the spatial and temporal coverage of nighttime OLI observations globally.

Land cover classification using Landsat 8 OLI in West Langsa Sub district, Langsa City

Land cover information is needed by various parties as a consideration in controlling land cover changes. The latest land cover information can be obtained using remote sensing techniques in the form of image classification maps. This technique is very effective in monitoring land cover because of its ability to quickly, precisely, and easily provide spatial information on the earth’s surface. The purpose of this study was to classify land cover in West Langsa Sub district, Langsa City using Landsat 8 OLI (Operational Land Imager) imagery. The classification method used in this study is the maximum likelihood classification (MLC) method. There are several considerations of various factors in the MLC method, including the probability of a pixel to be classified into a certain type or class. The results of Landsat 8 OLI image classification in West Langsa Sub district resulted in 6 land cover classes, namely mangrove forests, settlements, rice fields, shrubs, ponds and bodies of water. The largest land cover class is ponds with an area of 1981.54 ha (38.71%) and the smallest land cover is rice fields with an area of 115.58 ha (2.26%) of the total land cover class. Classification accuracy is indicated by the overall accuracy and kappa accuracy of 91.15% and 82.75%, respectively. These results meet the requirements set by the USGS (Overall Accuracy > 85%) and indicate that the Landsat 8 OLI image classification map can be used for various purposes.

Surface composition of debris-covered glaciers across the Himalaya using linear spectral unmixing of Landsat 8 OLI imagery

Abstract. The Himalaya mountain range is characterized by highly glacierized, complex, dynamic topography. The ablation area of Himalayan glaciers often features a highly heterogeneous debris mantle comprising ponds, steep and shallow slopes of various aspects, variable debris thickness, and exposed ice cliffs associated with differing ice ablation rates. Understanding the composition of the supraglacial debris cover is essential for a proper understanding of glacier hydrology and glacier-related hazards. Until recently, efforts to map debris-covered glaciers from remote sensing focused primarily on glacier extent rather than surface characteristics and relied on traditional whole-pixel image classification techniques. Spectral unmixing routines, rarely used for debris-covered glaciers, allow decomposition of a pixel into constituting materials, providing a more realistic representation of glacier surfaces. Here we use linear spectral unmixing of Landsat 8 Operational Land Imager (OLI) images (30 m) to obtain fractional abundance maps of the various supraglacial surfaces (debris material, clean ice, supraglacial ponds and vegetation) across the Himalaya around the year 2015. We focus on the debris-covered glacier extents as defined in the database of global distribution of supraglacial debris cover. The spectrally unmixed surfaces are subsequently classified to obtain maps of composition of debris-covered glaciers across sample regions. We test the unmixing approach in the Khumbu region of the central Himalaya, and we evaluate its performance for supraglacial ponds by comparison with independently mapped ponds from high-resolution Pléiades (2 m) and PlanetScope imagery (3 m) for sample glaciers in two other regions with differing topo-climatic conditions. Spectral unmixing applied over the entire Himalaya mountain range (a supraglacial debris cover area of 2254 km2) indicates that at the end of the ablation season, debris-covered glacier zones comprised 60.9 % light debris, 23.8 % dark debris, 5.6 % clean ice, 4.5 % supraglacial vegetation, 2.1 % supraglacial ponds, and small amounts of cloud cover (2 %), with 1.2 % unclassified areas. The spectral unmixing performed satisfactorily for the supraglacial pond and vegetation classes (an F score of ∼0.9 for both classes) and reasonably for the debris classes (F score of 0.7). Supraglacial ponds were more prevalent in the monsoon-influenced central-eastern Himalaya (up to 4 % of the debris-covered area) compared to the monsoon-dry transition zone (only 0.3 %) and in regions with lower glacier elevations. Climatic controls (higher average temperatures and more abundant precipitation), coupled with higher glacier thinning rates and lower average glacier velocities, further favour pond incidence and the development of supraglacial vegetation. With continued advances in satellite data and further method refinements, the approach presented here provides avenues towards achieving large-scale, repeated mapping of supraglacial features.

Landscape transformation after irrigation development in and around a semi-arid wetland ecosystem

The objectives of this study are to determine the spatial and temporal land use/cover changes in a semi-arid agricultural basin (Develi Basin) after the implementation of an irrigation project and to understand how these changes affected the wetlands (Sultan Marshes) located in the basin. The changes were determined using multitemporal Landsat Thematic Mapper and Landsat 8 Operational Land Imager imagery taken in 1987, 1998, 2007, and 2013. The images were classified into six information classes (grasslands/shrublands, croplands, permanent wetlands, water bodies, barren, and urban/built-up) using a hybrid classification method. Post-classification change detection was applied to determine the changes between different years. Overall, the accuracy of the classified images ranged from 85 to 94%. Grasslands/shrublands covered the largest area in the basin (63% in 2013), followed by croplands (32% in 2013). The area covered by water bodies, permanent wetlands, barren, and urban/built-up was 5% (in 2013). From 1987 to 2013, croplands expanded by 56%, while grasslands/shrublands declined by 15%. The areas occupied by water bodies decreased by 88% and permanent wetlands decreased by 4%. Urban/built-up areas expanded by 140%. The hydrologic regime of the Sultan Marshes wetland changed, which resulted in declines in water volumes by 85% and in water inflows by 55% from 2000 to 2015. Climatic variations during the 1987–2013 period were low and there was no apparent trend in precipitation and air temperature, which ruled out climatic conditions as one of the drivers of wetland changes. Economic and institutional factors supported the expansion of irrigated agriculture and animal husbandry in the basin and accelerated the expansion of croplands and conversion to industrial and fodder crops and orchards from traditional non-irrigated crops. Expansion of croplands and irrigated agriculture were the major drivers of the changes in the Sultan Marshes.

A Chlorophyll-a Algorithm for Landsat-8 Based on Mixture Density Networks

Retrieval of aquatic biogeochemical variables, such as the near-surface concentration of chlorophyll-a (Chla) in inland and coastal waters via remote observations, has long been regarded as a challenging task. This manuscript applies Mixture Density Networks (MDN) that use the visible spectral bands available by the Operational Land Imager (OLI) aboard Landsat-8 to estimate Chla. We utilize a database of co-located in situ radiometric and Chla measurements (N = 4,354), referred to as Type A data, to train and test an MDN model (MDNA). This algorithm’s performance, having been proven for other satellite missions, is further evaluated against other widely used machine learning models (e.g., support vector machines), as well as other domain-specific solutions (OC3), and shown to offer significant advancements in the field. Our performance assessment using a held-out test data set suggests that a 49% (median) accuracy with near-zero bias can be achieved via the MDNA model, offering improvements of 20 to 100% in retrievals with respect to other models. The sensitivity of the MDNA model and benchmarking methods to uncertainties from atmospheric correction (AC) methods, is further quantified through a semi-global matchup dataset (N = 3,337), referred to as Type B data. To tackle the increased uncertainties, alternative MDN models (MDNB) are developed through various features of the Type B data (e.g., Rayleigh-corrected reflectance spectra ρs). Using held-out data, along with spatial and temporal analyses, we demonstrate that these alternative models show promise in enhancing the retrieval accuracy adversely influenced by the AC process. Results lend support for the adoption of MDNB models for regional and potentially global processing of OLI imagery, until a more robust AC method is developed. Index Terms—Chlorophyll-a, coastal water, inland water, Landsat-8, machine learning, ocean color, aquatic remote sensing.

A machine learning approach to estimate chlorophyll-a from Landsat-8 measurements in inland lakes

Landsat-8 Operational Land Imager (OLI) provides an opportunity to map chlorophyll-a (Chla) in lake waters at spatial scales not feasible with ocean color missions. Although state-of-the-art algorithms to estimate Chla in lakes from satellite-borne sensors have improved, there are no robust and reliable algorithms to generate Chla time series from OLI imageries in turbid lakes due to the absence of a red-edge band and issues with atmospheric correction. Here, a machine learning approach termed the extreme gradient boosting tree (BST) was employed to develop an algorithm for Chla estimation from OLI in turbid lakes. This model was developed and validated by linking Rayleigh-corrected reflectance to near-synchronous in situ Chla data available from eight lakes in eastern China (N = 225) and three coastal and inland waters in SeaWiFS Bio-optical Archive and Storage System (N = 97). The BST model performed well on a subset of data (N = 102, R2 = 0.79, root mean squared difference = 7.1 μg L−1, mean absolute percentage error = 24%, mean absolute error = 1.4, Bias = 0.9), and had better Chla retrievals than several band-ratio algorithms and a random forest approach. The performance of BST model was judged as appropriate only for the range of conditions in the training data. Given these limitations, spatial and temporal variations of Chla in hundreds of lakes larger than 1 km2 in eastern China for the period of 2013–2018 were mapped using the BST model. OLI-derived Chla indicated that small lakes (<50 km2) had greater Chla than the larger lakes. This research suggests that machine-learning models provide practical approaches to estimate Chla in turbid lakes via broadband instruments like OLI and that extending to other regions requires training with a representative dataset.

Comparative analysis of mineral mapping for hyperspectral and multispectral imagery

The traditional approaches of mineral-mapping are time consuming and expensive process. Remote sensing is a tool to map the minerals precisely using their physical, chemical and optical properties. In the present study, Tirunelveli district in Tamil Nadu is selected to extract the abundant mineral such as Limestone using Hyperion and Landsat-8 OLI imageries. The chemical composition of the mineral is identified using scanning electron microscope (SEM) and energy dispersive X-ray spectroscopy (EDS) analysis. The spectral reflectance of minerals is characterized using analytical spectral device (ASD) field spectroradiometer. The minerals showed deep absorption in short wave infrared region from 1800 to 2500 nm. The mineral mapping in hyperspectral data is performed using various preliminary processing such as bad band removal, vertical strip removal, radiance and reflectance generation and postprocessing steps such as data dimensional reduction, endmember extraction and classification. To improve the classification accuracy, the vertical strip removal process is performed using a local destriping algorithm. Absolute reflectance of Hyperion and Landsat-8 OLI (Operational Land Imager) imageries is carried out using the FLAASH (fast line-of-sight atmospheric analysis of hypercubes) module. Spectral data reduction techniques in reflectance bands performed using minimum noise fraction method. The noiseless reflectance bands spatial data reduced by the Pixel Purity Index method in the threshold limit of 2.5 under 10,000 repetitions. The obtained reflectance imagery spectra compared with the spectral libraries such as USGS (United States Geological Survey), JPL (Jet Propulsion Laboratory) and field spectra. Endmembers of minerals are carried out using high probability score obtained from the various methods such as SAM (spectral angle mapper), SFF (spectral feature fitting) and BE (binary encoding). The mineral mapping of both imageries is carried out using a supervised classification approach. The results showed that hyperspectral remote sensing performed good results as compared to multispectral data.

Sea-Land Segmentation Using Deep Learning Techniques for Landsat-8 OLI Imagery

Automated coastline extraction from optical satellites is fundamental to coastal mapping, and sea-land segmentation is the core technology of coastline extraction. Deep convolutional neural networks (DCNNs) have performed well in semantic segmentation in recent years. However, sea-land segmentation using deep learning techniques remains a challenging task, due to the lack of a benchmark dataset and the difficulty of deciding which semantic segmentation model to use. We present a comparative framework of sea-land segmentation to Landsat-8 OLI imagery via semantic segmentation in deep learning techniques. Three issues are investigated: (1) constructing a sea-land benchmark dataset using Landsat-8 Operational Land Imager (OLI) imagery consisting of 18,000 km2 of coastline around China; (2) evaluating the feasibility and performance of sea-land segmentation by comparing the accuracy assessment, time complexity, spatial complexity and stability of state-of-the-art DCNNs methods; (3) choosing the most suitable semantic segmentation model for sea-land segmentation in accordance with Akaike information criterion (AIC) and Bayesian information criterion (BIC) model selection. Results show that the average test accuracy achieves over 99% accuracy, and the mean Intersection over Unions (mean IoU) is above 92%. These findings demonstrate that the Fully Convolutional DenseNet (FC-DenseNet) performs better than other state-of-the-art methods in sea-land segmentation, based on both AIC and BIC. Considering training time efficiency, DeeplabV3+ performs better for sea-land segmentation. The sea-land segmentation benchmark dataset is available at: https://pan.baidu.com/s/1BlnHiltOLbLKe4TG8lZ5xg.

Assessing the efficacy of Landsat-8 OLI imagery derived models for remotely estimating chlorophyll-a concentration in the Upper Ganga River, India

ABSTRACTChlorophyll-a (chl-a) serves as an indicator of productivity in surface water. Estimating chl-a concentration is pivotal for monitoring and subsequent conservation of surface water quality. Artificial neural network (ANN) based models were validated and tested for their efficacy against various regression models to determine the chl-a concentration in the Upper Ganga river. Landsat-8 Operational Land Imager (OLI) surface reflectance (SR) imagery for May and October along with in-situ data over a period of 2 years (2016–2017) was used to develop and validated models. Regression model performance was acceptable with a coefficient of determination (R2) of 0.57, 0.63, 0.66 and 0.68 for linear, exponential, logarithmic and power model, respectively. However, there was a significant improvement in the efficacy of chl-a determination using ANN model performance having a root mean square error (RMSE) of 1.52 µg l–1 and R2 = 0.97 in comparison to the best-performing regression model (power) with RMSE = 9.86 µg l–1 and R2 = 0.68. ANN exhibited comparatively more precise spatial and seasonal variability with mean absolute error (MAE) of 1.26 µg l–1 as compared to the best regression model (power) MAE = 7.98 µg l–1 suggesting the applicability of ANN for large-scale spatial and temporal monitoring river stretches using Landsat-8 OLI SR images.

Remote Monitoring of PSD Slope Under the Influence of Sand Dredging Activities in Lake Hongze Based on Landsat-8/OLI Data and VIIRS/DNB Night-Time Light Composite Data

Particle size distribution (PSD) is an important characterization parameter of the particle size structure of suspended particles, which is vulnerable to human activities such as sand dredging. In this study, sand dredging numbers were evaluated by the accumulated radiance algorithm using the night-time light composite data of the visible infrared imaging radiometer suite, and the PSD slope was derived by band ratio algorithm of Landsat-8/operational land imager (OLI) imagery. Then, the influence of sand mining on PSD slope was analyzed. We draw several crucial findings. First, the ratio of ${\boldsymbol{R}_{{{\bf rs}}({{{\bf Red}}})}}$ to ${\boldsymbol{R}_{{{\bf rs}}({{{\bf Green}}})}}$ based on OLI imagery was capable of retrieving PSD slope with good robustness ( R 2 = 0.71, p n = 41) and verified acceptance (mean absolute percentage error = 4.29%, root mean square error = 0.19, n = 20), and the PSD slope in HZL was lower in summer than in the other seasons from 2014 to 2017. Second, The night active vessels extracted from the night-time images show that sand dredging activities began at the outlet of Chengzi Lake (CZL) in April 2012, spread to the whole lake between 2014 and 2016, and then stopped in the northeast region in March 2017 due to the strict control of the local government. Third, the PSD slope is strongly influenced by sand mining, which is positively correlated with the number of sand dredgers ( R 2 = 0.92), i.e., the PSD slope increases 0.21 per 100 working sand dredgers in CZL.

Is the phenology-based algorithm for mapping deciduous rubber plantations applicable in an emerging region of northern Laos?

Updated information of rubber plantations is essential for assessing socioeconomic and environmental impacts, especially in the emerging region of northern tropics. Here, a phenological method was modified to detect rubber plantations using Landsat Operational Land Imager (OLI) imagery in Phongsaly Province of northern Laos, where it begun a rubber boom in the mid-2000s due to geo-economic cooperation. It highlighted the landscape and pixel differences of deciduous rubber plantations in the tri-temporal phases (i.e., pre-defoliation, defoliation, and foliation) during the dry season due to phenological changes. Six commonly used vegetation indices (VIs), including the Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EVI), Land Surface Water Index (LSWI), Atmospherically Resistant Vegetation Index (ARVI), Normalized Burn Ratio (NBR and NBR2) derived from OLI imagery during 2013–2016 were compared to determine the most suitable VI for discriminating the phenological differences of rubber plantations from natural forests. Then, the Differences of Normalized Burn Ratio (DNBR) was applied to generate the 30 m map of rubber plantations in 2016, by combining two masks of Landsat-derived forest and suitable elevation for rubber trees cultivation. The resultant map of rubber plantations had a classification accuracy of 93.7% and the Kappa coefficient of 0.848. Our study demonstrated the usefulness of the Landsat-derived tri-temporal phenological DNBR approach in an emerging region of northern Laos, despite requiring more scenes compared with single- and double temporal window methods.

A harmonized image processing workflow using Sentinel-2/MSI and Landsat-8/OLI for mapping water clarity in optically variable lake systems

This study demonstrates the applicability of harmonizing Sentinel-2 MultiSpectral Imager (MSI) and Landsat-8 Operational Land Imager (OLI) satellite imagery products to enable the monitoring of inland lake water clarity in the Google Earth Engine (GEE) environment. Processing steps include (1) atmospheric correction and masking of MSI and OLI imagery, and (2) generating scene-based water clarity maps in terms of Secchi depth (SD). We adopted ocean-color based atmospheric correction theory for MSI and OLI sensors modified with associated scene-specific metadata and auxiliary datasets available in GEE to generate uniform remote sensing reflectances (Rrs) products over optically variable freshwater lake surfaces. MSI-Rrs products derived from the atmospheric correction were used as input predictors in a bootstrap forest to determine significant band combinations to predict water clarity. A SD model for MSI (SDMSI) was then developed using a calibration dataset consisting of log-transformed SDin situ measurements (lnSDin situ) from 79 optically variable freshwater inland lakes collected within ±1 day of satellite overpass on 23-Aug 2017 (MAE = 0.53 m) and validated with 276 samples collected within ±1 day of a 12-Sep 2017 image (MAE = 0.66 m) across three ecoregions in Minnesota, USA. A separate SD model for MSI was also developed using similar spectral bands present on the OLI sensor (SDsOLI) where cross-sensor performance can be evaluated during coincident overpass events. SDsOLI applied to both MSI and OLI (SDOLI) models were further validated using two coincident overpass sets of imagery on 27-Sep 2017 (n = 18) and 13-Aug 2018 (n = 43), yielding a range of error from 0.25 to 0.67 m. Potential sources of model errors and limitations are discussed. Data derived from this multi-sensor methodology is anticipated to be used by researchers, lake resource managers, and citizens to expedite the pre-processing steps so that actionable information can be retrieved for decision making.

A dryland cover state mapping using catastrophe model in a spectral endmember space of OLI: a case study in Minqin, China

ABSTRACTAlternative state theory with catastrophe conceptual model is very attractive for dryland zero net land degradation (ZNLD) management, but practically challenging for mapping cover state and pathway with conventional approaches. This paper developed a methodological framework in the generic spectral endmember space (SVD, soils and substrates, vegetation, and dark surface endmembers ) of Landsat-8 OLI (operational land imager) imagery to concretize the catastrophe model of dryland cover state based on the co-development of plant and soil variables in a landscape. A cover function index (LCI) was proposed together with cover type growth form as cover variables in multi-seasonal SVD space. Furthermore, ecological site (ES) and topographic wetness index (TWI) were used as the landscape variables. The four variables as the explanatory variates, soil organic matter (SOM) as a slow response variable of dryland cover state, were explored by classification and regression trees (CART) for the state space. Using the Minqin dryland system in China as a case study, CART explored 65.9% amounts of SOM variance of 85 training samples with the four explanatory variables, and the measured SOM had liner relationship with CART predicted SOM at 0.05 significant level with the explained 39% SOM variance for 10 test samples. In the learned transparent tree, the two alternative stable states were identified correctly and validated by another soil intrinsic clay content variable independently, one was the current highest average SOM (26.757 g kg−1) of perennial forest/shrub covers, the other was the bared or sparse vegetation cover with low average SOM (≤8.518 g kg−1) as a critical SOM limit in a given ecological site type. The mean clay content of the undesirable alternative states was all below 60 g kg−1 with severe soil physical degradation, and had a clear breaking gap with other states. Moreover, the SOM-LCI plane projected the unstable states being folded in the basins of the attraction of two alternative stable states (i.e. 0.39–0.76 LCI) with relative resilience analysis of pathways. In the 10 test sites, only 1 site of the state was not predicted correctly. The results suggest the model has a level of predictability for dryland cover state and its pathway. According to the estimated states, there has 12.74% study area improved and continued to recovering the desirable states, and the study area has already net improvement at the landscape level.Therefore, without the comparable series of historical data for adaptive management, the developed framework in SVD of Landsat imagery is a cost-effective and time-saving tool, which helps managers to assess and develop appropriate management with a more spatially diversified strategy for zero net land degradation using a space-for-time substitution in an era of unprecedented change.

Geospatial hazard modeling for the delineation of flash flood-prone zones in Wadi Dahab basin, Egypt

Abstract This study delineated flash flood hazard zones leading to vast destruction to infrastructure, property, and loss of life. An integrated approach using remote sensing and geographic information systems was applied to reveal flash flood-prone zones. The study approach evaluated topographic, geologic, and hydrologic factors holistically to assess these hazard zones. The morphometric characteristics of Wadi Dahab sub-basins were supported by topographic, geologic, and hydrologic information. Data from Shuttle Radar Terrain Mission and Operational Land Imager imagery were analyzed to characterize hydrological morphometrics, lithology, soil types, and land use. A Natural Resources Conservation Service model was selected to calculate runoff depth at ungauged watersheds. A spatially distributed unit hydrograph was adopted to create the flow time and runoff velocity. The Flashflood Hazard Model was developed by spatial integration of all contributing factors. An analytical hierarchy process was adopted for the logic ranking of the effective factors. The flash flood hazard map classifies Wadi Dahab basin into five relative hazard zones: very high, high, moderate, low, and very low. The highly hazardous zones are distributed at the downstream of Wadi Dahab basin corresponding to steep topography and Precambrian rocks. The hazard map was validated using the flash flood markers defined from field observations.

Modification of 6SV to remove skylight reflected at the air-water interface: Application to atmospheric correction of Landsat 8 OLI imagery in inland waters.

During the atmospheric correction of remote sensing data in inland waters, the original Second Simulation of the Satellite Signal in the Solar Spectrum-Vector version (6SV) model does not eliminate the specular reflection of downward skylight radiance at the air-water interface. Thus, we propose a modified version of the 6SV model (M6SV) that does remove reflected skylight at the air-water interface. We apply the new model to the atmospheric correction of a Landsat 8 Operational Land Imager (OLI) image over Taihu Lake, China, where the aerosol optical depth is known. In situ reflectance measurements acquired concurrently with the L8/OLI image are used to validate the performance of the new M6SV algorithm. To further analyze the merits and demerits of M6SV, the model is compared with two short-wave infrared (SWIR)-based atmospheric correction models: the Sea-Viewing Wide Field-of-View Sensor Data Analysis System short-wave infrared (SD-SWIR) model and the Vanhellemont & Ruddick short-wave infrared with a per scene fixed aerosol type (VR-SWIR-F) model. Comparisons of results from all three L8/OLI image atmospheric corrections with the in situ remote sensing reflectance data show that M6SV produces reliable atmospheric corrections in the green and red spectral bands and is an effective alternative for Landsat 8 OLI atmospheric correction in inland waters.

A Unifying Approach to Classifying Wetlands in the Ontonagon River Basin, Michigan, Using Multi-temporal Landsat-8 OLI Imagery

Accurate spatial information is critical to the assessment and protection of wetlands in the context of human intervention and global climate change. However, it is challenging to map and monitor wetland vegetation classes with satisfactory results because of their highly seasonal dynamics, spatial heterogeneity, and spectral similarity. This paper examines the effectiveness of various classification approaches commonly employed in wetland mapping, including the support vector machine (SVM) algorithm, maximum likelihood classifier (MLC), classification and regression tree (CART) and other remote sensing indices, by using multi-temporal Landsat-8 Operational Land Imager (OLI) spectral data and end-member fraction data as well as terrain data. These different mapping approaches were compared in the Ontonagon River drainage basin in upper Michigan, USA, where easily-confused wetland types such as forested wetland, palustrine scrub/shrub wetland, and palustrine emergent wetland are extensively distributed. The results show that multi-temporal data sets can reduce the classification omission caused by the highly seasonal dynamic of wetlands. The spatial heterogeneity can partly be characterized by using end-member fraction maps. Classification of the fraction maps had better results than that using the original spectral data, which implies that the selection of data inputs could be more important than the selection of classifiers. While each algorithm has its own capability of discriminating specific wetland types, CART showed relatively better classification results than the others due to its compatibility and lack of assumptions about data normal distribution. Finally, the authors developed a decision tree method (called DTM), which adopted the satisfactory resultant classification of specific wetland types using MLC and SVM classification, to update the wetland map of Ontonagon River Basin with acceptable accuracy (overall accuracy of 89% and kappa coefficients of 0.89).

Lithological Classification Using Sentinel-2A Data in the Shibanjing Ophiolite Complex in Inner Mongolia, China

As a source of data continuity between Landsat and SPOT, Sentinel-2 is an Earth observation mission developed by the European Space Agency (ESA), which acquires 13 bands in the visible and near-infrared (VNIR) to shortwave infrared (SWIR) range. In this study, a Sentinel-2A imager was utilized to assess its ability to perform lithological classification in the Shibanjing ophiolite complex in Inner Mongolia, China. Five conventional machine learning methods, including artificial neural network (ANN), k-nearest neighbor (k-NN), maximum likelihood classification (MLC), random forest classifier (RFC), and support vector machine (SVM), were compared in order to find an optimal classifier for lithological mapping. The experiment revealed that the MLC method offered the highest overall accuracy. After that, Sentinel-2A image was compared with common multispectral data ASTER and Landsat-8 OLI (operational land imager) for lithological mapping using the MLC method. The comparison results showed that the Sentinel-2A imagery yielded a classification accuracy of 74.5%, which was 2.5% and 5.08% higher than those of the ASTER and OLI imagery, respectively, indicating that Sentinel-2A imagery is adequate for lithological discrimination, due to its high spectral resolution in the VNIR to SWIR range. Moreover, different data combinations of Sentinel-2A + ASTER + DEM (digital elevation model) and OLI + ASTER + DEM data were tested on lithological mapping using the MLC method. The best mapping result was obtained from Sentinel-2A + ASTER + DEM dataset, demonstrating that OLI can be replaced by Sentinel-2A, which, when combined with ASTER, can achieve sufficient bandpasses for lithological classification.

Mapping Soil Salinity/Sodicity by using Landsat OLI Imagery and PLSR Algorithm over Semiarid West Jilin Province, China

Soil salinity and sodicity can significantly reduce the value and the productivity of affected lands, posing degradation, and threats to sustainable development of natural resources on earth. This research attempted to map soil salinity/sodicity via disentangling the relationships between Landsat 8 Operational Land Imager (OLI) imagery and in-situ measurements (EC, pH) over the west Jilin of China. We established the retrieval models for soil salinity and sodicity using Partial Least Square Regression (PLSR). Spatial distribution of the soils that were subjected to hybridized salinity and sodicity (HSS) was obtained by overlay analysis using maps of soil salinity and sodicity in geographical information system (GIS) environment. We analyzed the severity and occurring sizes of soil salinity, sodicity, and HSS with regard to specified soil types and land cover. Results indicated that the models’ accuracy was improved by combining the reflectance bands and spectral indices that were mathematically transformed. Therefore, our results stipulated that the OLI imagery and PLSR method applied to mapping soil salinity and sodicity in the region. The mapping results revealed that the areas of soil salinity, sodicity, and HSS were 1.61 × 106 hm2, 1.46 × 106 hm2, and 1.36 × 106 hm2, respectively. Also, the occurring area of moderate and intensive sodicity was larger than that of salinity. This research may underpin efficiently mapping regional salinity/sodicity occurrences, understanding the linkages between spectral reflectance and ground measurements of soil salinity and sodicity, and provide tools for soil salinity monitoring and the sustainable utilization of land resources.

Spatio-temporal differences in cloud cover of Landsat-8 OLI observations across China during 2013–2016

Currently, the historical archive images of Landsat family sensors are probably the most effective data products for tracking global longitudinal changes since the 1970s. However, the issue of the degree and extent of cloud coverage is always a challenge and varies distinctively worldwide. So far, acquisition probability (AP) analyses of cloud cover (CC) of Landsat observations have been conducted with different sensors at regional scale. To our knowledge, CC probability analysis for the newly-launched Landsat-8 Operational Land Imager (OLI) across China is not reported. In this paper, monthly, seasonal, and annual APs for Landsat OLI (44,228 in total) images over China acquired from April 2013 to October 2016 with various CC thresholds were analyzed. The results showed that: first, the cumulative average APs of all OLI data over China at the CC thresholds ≤30% was about 49.6% which illustrated the availability of OLI imagery across China. Second, the spatial patterns of 10%, 20%, and 30% CC thresholds of OLI observations, coincided well with the precipitation distributions separated by the respective 200 mm, 400 mm, and 800 mm isohyetal lines. Third, the APs of images with the 30% CC threshold are the highest in autumn and winter especially in October of 58.7%, while the corresponding lowest probability occurred in June of 41.0%. Finally, the spatial differences in APs of targeted images with ≤30% CC thresholds were quite significant. At regional scales, the arid and semi-arid areas, Inland River and Songliao River basins, and northwestern side of the Hu Huanyong population line had the larger probabilities of obtaining high-quality images. Our study suggested that OLI imagery satisfy the data requirements needed for land surface monitoring, although there existed obvious spatio-temporal differences in APs over China at the 30% CC threshold.

A robust Multi-Band Water Index (MBWI) for automated extraction of surface water from Landsat 8 OLI imagery

Surface water is vital resources for terrestrial life, while the rapid development of urbanization results in diverse changes in sizes, amounts, and quality of surface water. To accurately extract surface water from remote sensing imagery is very important for water environment conservations and water resource management. In this study, a new Multi-Band Water Index (MBWI) for Landsat 8 Operational Land Imager (OLI) images is proposed by maximizing the spectral difference between water and non-water surfaces using pure pixels. Based on the MBWI map, the K-means cluster method is applied to automatically extract surface water. The performance of MBWI is validated and compared with six widely used water indices in 29 sites of China. Results show that our proposed MBWI performs best with the highest accuracy in 26 out of the 29 test sites. Compared with other water indices, the MBWI results in lower mean water total errors by a range of 9.31%–25.99%, and higher mean overall accuracies and kappa coefficients by 0.87%–3.73% and 0.06–0.18, respectively. It is also demonstrated for MBWI in terms of robustly discriminating surface water from confused backgrounds that are usually sources of surface water extraction errors, e.g., mountainous shadows and dark built-up areas. In addition, the new index is validated to be able to mitigate the seasonal and daily influences resulting from the variations of the solar condition. MBWI holds the potential to be a useful surface water extraction technology for water resource studies and applications.