RSWetlands

Using a #Landsat time series and #objectbased and #decisiontree classification, Mao et al. examine #wetland change in the #AmurRiverBasin, which crosses China, Russia, and Mongolia, noting that drivers of change differ across countries. #rswetlands sciencedirect.com/science/articl…

@chloebrown_RS, @boyd_doreen, @hugoloper1

Using a combination of field data, #Sentinel2 imagery, and linear regression, Sjögersten et al. quantify #carbon stocks of #wetlands in #Mexico, concluding that swamp and marsh store substantial amounts of carbon but lack formal protection. #rswetlands sciencedirect.com/science/articl…

Thamaga et al. review #remotesensing applications for the purpose of mapping #wetlands in #SubsaharanAfrica, emphasizing the importance of cheap and readily available data and techniques and the use of #machinelearning algorithms. #rswetlands tandfonline.com/doi/full/10.10…

Using #Landsat8 to estimate land surface temperature, Xue et al. examine the urban heat island effect and the cooling impact of #wetlands, finding that wetlands exhibit stronger cooling than #greenspaces, particularly those with complex shapes #rswetlands sciencedirect.com/science/articl…

Ghosh and Behera compare a semi-empirical and a deep learning model for estimating #mangrove AGB using #Sentinel1 backscatter and coherence data, concluding that #deeplearning better captures the relationship between S1 variables and AGB. #rswetlands sciencedirect.com/science/articl…

Pena-Regueiro et al. use #Sentinel2 #NDWI to capture and assess changes to the extent of small water bodies within four #wetlands along the coast of the #MediterraneanSea which display diverse temporal and spatial flooding patterns. #rswetlands mdpi.com/2073-4441/12/5….

Using #Sentinel1 data to discriminate water boundaries and soil roughness and #Landsat8 and #Sentinel2 data to monitor soil moisture, Hakdaoui et al. develop a change detection technique to monitor desert #wetlands after a flash-flood event. #rswetlands mdpi.com/2072-4292/11/9…

@SahebaBhatnagar, @BidishaQ

Bhatnagar et al. evaluate the ability of #Sentinel2 to classify vegetation communities in wetlands for the purpose of monitoring ecological health. Using a #baggedtree classifier, vegetation communities are classified with accuracies ≤ 87%. #rswetlands sciencedirect.com/science/articl…

@PaulieJTails, @DLagomasino, @EarthToLola, @carBenPoulter

In the #NorthAtlanticBasin, Taillie et al. observe 30x more #mangrove damage during the 2017 Atlantic Hurricane season relative to previous seasons using #NDVI data, a cause for concern given potential increases in mega hurricane frequency. #rswetlands iopscience.iop.org/article/10.108…

Forested depressional #wetland inundation is difficult to capture using traditional #remotesensing methods. Du et al. develop a #deeplearning network capable of capturing inundation using #WV3 and topographic data with an accuracy of 95%. #rswetlands mdpi.com/2072-4292/12/4…

Simioni et al. examine 4 #machinelearning algorithms and active/passive #remotesensing features to determine which is best delineate inland #marsh #wetlands. While k-NN produced the highest OA, feature importance is dependent on data inputs. #rswetlands link.springer.com/article/10.100…

To improve #wetland monitoring, Kaplan and Avdan investigate the potential correlation between parameters derived from #Landsat8 and #Sentinel1, finding, for example, a correlation between optical and thermal parameters and VH backscatter. #rswetlands mdpi.com/2220-9964/7/10…

@aymhamid, @IslamAbouElMag1

Abdel-Hamid et al. find that integration of ALOS/PALSAR L-band, #RapidEye optical imagery, #OBIA, and random forest #machinelearning is best for mapping #mangrove extent along the #RedSea coastline in Egypt with an overall accuracy of ~92%. #rswetlands mdpi.com/2071-1050/10/3…