A highly automated algorithm for wetland detection using multi-temporal optical satellite data

Given global climate change and rapid land cover changes due to human activities, accurately identifying, extracting, and monitoring the long-term evolution of wetland resources is profoundly significant, particularly in areas with fragile ecological conditions. Gansu Province, located in northwest China, contains all wetland types except coastal wetlands. The complexity of its wetland types has resulted in a lack of accurate and comprehensive information on wetland changes. Using Gansu Province as a case study, we employed the GEE platform and Landsat time-series satellite data, combining high-quality sample datasets with feature-optimized multi-source feature sets. The random forest algorithm was utilized to create wetland classification maps for Gansu Province across eight periods from 1987 to 2020 at a 30 m resolution and to quantify changes in wetland area and type. The results showed that the wetland mapping method achieved robust classification results, with an average overall accuracy (OA) of 96.0% and a kappa coefficient of 0.954 across all years. The marsh type exhibited the highest average user accuracy (UA) and producer accuracy (PA), at 96.4% and 95.2%, respectively. Multi-source feature aggregation and feature optimization effectively improve classification accuracy. Topographic and seasonal features were identified as the most important for wetland extraction, while textural features were the least important. By 2020, the total wetland area in Gansu Province was 10,575.49 km2, a decrease of 4536.86 km2 compared to 1987. The area of marshes decreased the most, primarily converting into grasslands and forests. River, lake, and constructed wetland types generally exhibited an increasing trend with fluctuations. This study provides technical support for wetland ecological protection in Gansu Province and offers a reference for wetland mapping, monitoring, and sustainable development in arid and semi-arid regions.

Performance evaluation of object based greenhouse detection from Sentinel-2 MSI and Landsat 8 OLI data: A case study from Almería (Spain)

Assessment of Earth Observations Data Harmonization

Sentinel-2 MSI (MultiSpectral Instrument) is the Landsat-like spatial resolution (10–60 m) super-spectral instrument of the European Space Agency (ESA), aimed at additional data continuity for global land surface monitoring with Landsat and Satellite Pour l'Observation de la Terre (SPOT) missions. Several studies with simulated and real data have been conducted in the last several years to show the potential of Sentinel-2 MSI, including its use for geological remote sensing. An aspect of Sentinel-2 MSI that has not yet been evaluated for geological remote sensing is the 5–10 days revisiting time. This paper shows the first results of a multi-temporal study performed in a hydro-thermal alteration system in southern Spain. Several band ratios for mapping mineral alteration have been calculated in more than 200 Sentinel-2 MSI images that cover a period of two years. Results show the effect of seasonality, illumination and weather on ground cover and the stability of the geological indices.

Iron is the fourth most common mineral found in the earth crust. Although it may not be as important for soil fertility as, e.g., phosphorus, nitrogen and organic matter, its absence would be detrimental to plant growth. At the same time, iron oxides are highly correlated with phosphorus availability. Iron is thus an indicator for soil fertility and the usability of an area for cultivation of crop. A relatively high spectral resolution is needed for mapping iron oxide contents with spectral reflectance data, and remote sensing is the only suitable tool for surveying large areas at a high temporal and spatial interval. Sentinel-2 MSI (MultiSpectral Instrument) is the Landsat-like spatial resolution (10–60 m) super-spectral instrument of the European Space Agency (ESA), aimed at additional data continuity for global land surface monitoring with Landsat and Satellite Pour l’Observation de la Terre (SPOT) missions. Several studies with simulated and real data have been conducted in the last several years to show the potential of Sentinel-2 MSI, including its use for geological remote sensing, mineral mapping in particular. Sentinel-2 has several bands that cover the 0.9 μm iron absorption feature, while space-borne sensors traditionally used for geologic remote sensing, like ASTER and Landsat, had only one band in this feature. In this paper, we show a comparison of Sentinel-2 and AVIRIS to demonstrate the usability of the VNIR bands for mapping the near-infrared iron absorption feature. Next, we present spectral indices for mapping iron minerals that are important in soil fertility and mineral exploration.

The domain of high-resolution and very high-resolution optical sensors and Synthetic Aperture Radars (SAR) has dramatically increased. Dependable satellites providing accurate and reliable measurements are key to essential urgent activities such as climate change monitoring. As Earth Observation (EO) system architect in Europe, the European Space Agency (ESA) needs to assess the quality and the suitability of the EO data provided by its own satellites and to promote awareness about the importance of assessing EO data quality provided by other missions, e.g. in the commercial and New Space sector. Also, this would allow establishing a dialog with the EO data mission providers in order to improve the overall coherence of the EO system and data interoperability. Furthermore, some of these missions are potential candidates to become ESA Third Party Missions (TPM), pending ESA selection. Calibration/Validation (Cal/Val) activities are fundamental to assess the quality of EO satellite data, requiring dedicated measurements taken from established reference Cal/Val sites and networks. However, Cal/Val sites are often customized to the operator's needs, with little to no flexibility. Therefore, ESA is currently willing to explore the new idea of an open and pioneering reference site, named as Cal/Val Park. The concept is currently under investigation and is meant to be a multi-Agency cooperation.

As one of the world's largest freshwater ecosystems, the Great Lakes Basin houses thousands of acres of wetlands that support a variety of crucial ecological and environmental functions at the local, regional, and global scales. Monitoring these wetlands is critical to conservation and restoration efforts; however, current methods that rely on field monitoring are labor-intensive, costly, and often outdated. In this article, we present a graphical user interface constructed in Google Earth Engine called the Wetland Extent Tool (WET), which allows semiautomatic wetland classification according to a user-input area of interest and date range. WET conducts multisource, moderate resolution processing utilizing Landsat 8 Operational Land Imager, Sentinel-2 MultiSpectral Instrument, Sentinel-1 C-SAR, and Shuttle Radar Topography Mission (SRTM) datasets to classify wetlands in the entire Great Lakes Basin. We evaluated classification results of wetlands, uplands, and open water from May–September 2019, and tested whether SRTM elevation, slope, or the Dynamic Surface Water Extent produced the most accurate results in each Great Lake Basin in conjunction with optical indices and radar composites. We found that slope produced the most accurate classification in Lake Michigan, Huron, Superior, and Ontario, while elevation performed best in Lake Erie. Classification results averaged 86.2% overall accuracy, 70.0% wetland consumer's accuracy, and 82.7% wetland producer's accuracy across the Great Lakes Basin. WET leverages cloud-computing for multisource processing of moderate resolution remote sensing data, and employs a user interface in Google Earth Engine that wetland managers and conservationists can use to monitor wetland extent in the Great Lakes Basin in near real-time.

Coastal wetland mapping plays an essential role in monitoring climate change, the hydrological cycle, and water resources. In this study, a novel classification framework based on the gravitational optimized multilayer perceptron classifier and extended multi-attribute profiles (EMAPs) is presented for coastal wetland mapping using Sentinel-2 multispectral instrument (MSI) imagery. In the proposed method, the morphological attribute profiles (APs) are firstly extracted using four attribute filters based on the characteristics of wetlands in each band from Sentinel-2 imagery. These APs form a set of EMAPs which comprehensively represent the irregular wetland objects in multiscale and multilevel. The EMAPs and original spectral features are then classified with a new multilayer perceptron (MLP) classifier whose parameters are optimized by a stability-constrained adaptive alpha for a gravitational search algorithm. The performance of the proposed method was investigated using Sentinel-2 MSI images of two coastal wetlands, i.e., the Jiaozhou Bay and the Yellow River Delta in Shandong province of eastern China. Comparisons with four other classifiers through visual inspection and quantitative evaluation verified the superiority of the proposed method. Furthermore, the effectiveness of different APs in EMAPs were also validated. By combining the developed EMAPs features and novel MLP classifier, complicated wetland types with high within-class variability and low between-class disparity were effectively discriminated. The superior performance of the proposed framework makes it available and preferable for the mapping of complicated coastal wetlands using Sentinel-2 data and other similar optical imagery.

Knowledge gained about mangrove species mapping is essential to understand mangrove species’ development and to better estimate their ecological service value. Spectral bands and spatial resolution of remote sensing data are two important factors for accurate discrimination of mangrove species. Mangrove species classification in Shenzhen Bay, China, was performed using Sentinel-2 (S2) multispectral instrument (MSI) data and Google Earth (GE) high-resolution imagery as data sources, and their suitability in mapping mangrove forest at a species level was examined. In the classification feature groups, the spectral bands were from the S2 MSI data and the textural features were based on GE imagery. The support vector machine classifier was used in mangrove species classification processing with eight groups of features, which were based on different S2 spectral bands and different GE spatial resolution textural features. The highest overall accuracy of our mapping results was 78.57% and the kappa coefficient was 0.74, which indicated great potential for using the combination of S2 MSI and GE imagery for distinguishing and mapping mangrove species.

An up-to-date wetlands map based on remote sensing data at a continental scale is urgently needed for estimating global environmental change. In this study, a wetlands map of North America was developed using Moderate Resolution Imaging Spectroradiometer (MODIS) data obtained in 2008 and ancillary data. For this purpose, a decision rule classification method was developed relied upon the hierarchical characteristics of land types and prior knowledge about the geographical location of wetlands. Two hierarchical levels of land types were used to extract wetlands. At the first level, non-vegetation land types including water, snow, urban, and bare areas were separately extracted from vegetation land types using threshold methods. At the second level, wetlands were discriminated from non-wetland vegetation land types with the MODIS tasseled cap (brightness, greenness, and wetness) indices using the decision tree method. In addition, elevation data were used to build the elevation mask and a climate map was used to subdivide the study area into five sub-regions. In the quantitative accuracy assessment, user's and producer's accuracies of wetlands for the whole study area were calculated as 80.3% and 83.7%, respectively. In a comparison with two existing global land cover datasets, GLC2000 and IGBP DISCover, our results show significant improvement in extracting coastal and narrow types of wetlands. This study indicates that decision rule classification, integrated with multi-temporal MODIS data and ancillary data, is useful to develop an improved wetlands map at a continental scale.

Land use and land cover (LULC) mapping initiatives are essential to support decision making related to the implementation of different policies. There is a need for timely and accurate LULC maps. However, building them is challenging. LULC changes affect natural areas and local biodiversity. When they cause landscape fragmentation, the mapping and monitoring of changes are affected. Due to this situation, improving the efforts for LULC mapping and monitoring in fragmented biomes and ecosystems is crucial, and the adequate separability of classes is a key factor in this process. We believe that combining multidimensional Earth observation (EO) data cubes and spectral vegetation indices (VIs) derived from the red edge, near-infrared, and shortwave infrared bands provided by the Sentinel-2/MultiSpectral Instrument (S2/MSI) mission reduces uncertainties in area estimation, leading toward more automated mappings. Here, we present a low-cost semi-automated classification scheme created to identify croplands, pasturelands, natural grasslands, and shrublands from EO data cubes and the Surface Reflectance to Vegetation Indexes (sr2vgi) tool to automate spectral index calculation, with both produced in the scope of the Brazil Data Cube (BDC) project. We used this combination of data and tools to improve LULC mapping in the Brazilian Cerrado biome during the 2018–2019 crop season. The overall accuracy (OA) of our results is 88%, indicating the potential of the proposed approach to provide timely and accurate LULC mapping from the detection of different vegetation patterns in time series.

With a five-day revisit frequency over coastal regions and a spatial resolution of 10–60 m, the Sentinel-2 multispectral instrument (MSI) has shown its capacity to provide a reasonably accurate remote sensing reflectance ( $R_{\mathrm {rs}}$ ) data product over water when the standard “black pixel” (BP) atmospheric correction algorithm was applied to the top-of-atmospheric (TOA) reflectance data. Alternative atmospheric correction approaches, such as the POLYnomial-based algorithm applied to Medium Resolution Imaging Spectrometer (MERIS) (POLYMER), may show advantages under nonoptimal observation conditions (e.g., in the presence of strong sun glint). Here, POLYMER is implemented to process the data collected by both MSI and the Moderate Resolution Imaging Spectroradiometer (MODIS) with the resulting $R_{\mathrm {rs}}$ evaluated with concurrent and colocated in situ $R_{\mathrm {rs}}$ data collected from the AERONET-OC platforms. The results indicate less uncertainties in the MSI $R_{\mathrm {rs}}$ than those in the MODIS $R_{\mathrm {rs}}$ , and also less uncertainties in the MSI $R_{\mathrm {rs}}$ than those reported earlier. This is possibly attributed to the spatial heterogeneity of coastal waters where MODIS coarse-resolution data may suffer, and to the high-quality AERONET-OC data. In addition, for the evaluation data set, MSI $R_{\mathrm {rs}}$ does not appear to suffer from adjacency effects from the AERONET-OC platform and clouds, leading to more coverage than MODIS in nearshore waters. However, MSI $R_{\mathrm {rs}}$ is noisy in relatively clear waters, possibly due to the noisy TOA reflectance in the atmospheric correction bands over clear waters.

Inland and coastal waters provide key ecosystem services and are closely linked to human well-being. In this study, we propose a semi-analytical method, which can be applied to Sentinel-2 MultiSpectral Instrument (MSI) images to retrieve high spatial-resolution total suspended solids (TSS) concentration in a broad spectrum of aquatic ecosystems ranging from clear to extremely turbid waters. The presented approach has four main steps. First, the remote sensing reflectance (Rrs) at a band lacking in MSI (620 nm) is estimated through an empirical relationship from Rrs at 665 nm. Second, waters are classified into four types (clear, moderately turbid, highly turbid, and extremely turbid). Third, semi-analytical algorithms are used to estimate the particulate backscattering coefficient (bbp) at a reference band depending on the water types. Last, TSS is estimated from bbp at the reference band. Validation and comparison of the proposed method with three existing methods are performed using a simulated dataset (N = 1000), an in situ dataset collected from global inland and coastal waters (N = 1265) and satellite matchups (N = 40). Results indicate that the proposed method can improve TSS estimation and provide accurate retrievals of TSS from all three datasets, with a median absolute percentage error (MAPE) of 14.88 %, 31.50 % and 41.69 % respectively. We also present comparisons of TSS mapping between the Sentinel-3 Ocean and Land Colour Instrument (OLCI) and MSI in Lake Kasumigaura, Japan and the Tagus Estuary, Portugal. Results clearly demonstrate the advantages of using MSI for TSS monitoring in small water bodies such as rivers, river mouths and other nearshore waters. MSI can provide more detailed and realistic TSS estimates than OLCI in these water bodies. The proposed TSS estimation method was applied to MSI images to produce TSS time-series in Lake Kasumigaura, which showed good agreements with in situ and OLCI-derived TSS time-series.

We investigated the predictive strength of forested wetland maps produced using digital elevation models (DEMs) derived from Light Detection and Ranging (LiDAR) data and multiple topographic metrics, including multiple topographic wetness indices (TWIs), a TWI enhanced to incorporate information on water outlets, normalized relief, and hybrid TWI/relief in the Coastal Plain of Maryland. LiDAR DEM based wetland maps were compared to maps of inundation and existing wetland maps. TWIs based on the most distributed FD8 (8 cells) and somewhat distributed D∞ (1–2 cells) flow routing algorithms were better correlated with inundation than a TWI based on a non-distributed D8 (1 cell) flow routing algorithm, but D∞ TWI class boundaries appeared artificial. The enhanced FD8 TWI provided good prediction of wetland location but could not predict periodicity of inundation. Normalized relief provided good prediction of inundation periodicity but was less able to map wetland boundaries. A hybrid of these metrics provided good measurement of wetland location and inundation periodicity. Wetland maps based on topographic metrics included areas of flooded forest that were similar to an aerial photography based wetland map. These results indicate that LiDAR based topographic metrics have potential to improve accuracy and automation of wetland mapping.

ABSTRACTThis study presents the first comparison of Landsat 8 Operational Land Imager (OLI) and Sentinel-2 Multispectral Instrument (MSI) in identifying soil salinity using soil physiochemical, spectral, statistical, and image analysis techniques. By the end of the century, intermediate sea level rise scenarios project approximately 1.3 meters of sea level rise along the coast of the southeastern United States. One of the most vulnerable areas is Hyde County, North Carolina, where 1140 km2 of agricultural lands are being salinized, endangering 4,200 people and $40 million USD of property. To determine the best multispectral sensor to map the extent of salinization, this study compared the feasibility of OLI and MSI to estimate electrical conductivity (EC). The EC of field samples were correlated with handheld spectrometer spectra resampled into multispectral sensor bands. Using an iterative ordinary least squares regression, it was found that EC was sensitive to OLI bands 2 (452 nm – 512 nm) and 4 (636 nm – 673 nm) and MSI bands 2 (457.5 nm – 522.5 nm) and 4 (650 nm – 680 nm). Respectively, the R2Adj and Root Mean Square Error (RMSE) of 0.04–0.54 and 1.15 for OLI, and 0.05–0.67 and 1.17 for MSI, suggests that the two sensors have similar salinity modelling skill. The extracted saline soils make up approximately 1,703 hectares for OLI and 118 hectares for MSI, indicating overestimation from the OLI image due to its coarser spatial resolution. Additionally, field samples indicate that nearby vegetated land is saline, indicating an underestimation of total impacted land. As sea levels rise, accurately monitoring soil salinization will be critical to protecting coastal agricultural lands. MSI’s spatial and temporal resolution makes it superior to OLI for salinity tracking though they have roughly equivalent spectral resolutions. This study demonstrates that visible spectral bands are sensitive to soil salinity with the Blue and Red spectral ranges producing the highest model accuracy; however, the low accuracies for both sensors indicate the need of narrowband sensors. The HyspIRI to be launched in the early 2020s by NASA may provide ideal data source in soil salinity studies.