Please use this identifier to cite or link to this item: https://repositorio.uca.edu.ar/handle/123456789/9005
Título : Flooded-area satellite monitoring within a Ramsar wetland Nature Reserve in Argentina
Autor : Ferral, Anabella 
Luccini, Eduardo 
Aleksinkó, Alejandro 
Scavuzzo, Carlos M. 
Palabras clave : HUMEDALESBIODIVERSIDADAGUAECOSISTEMARESERVAS NATURALESMEDIO AMBIENTE
Fecha de publicación : 2019
Editorial : Elsevier
Cita : Ferral, A. Luccini, E. Aleksinkó, A. Scavuzzo, Carlos M. Flooded-area satellite monitoring within a Ramsar wetland Nature Reserve in Argentina [en línea]. Postprint del artículo publicado en Remote Sensing Applications: Society and Environment. 2019, 15. doy. 10.1016/j.rsase.2019.04.003. Disponible en: https://repositorio.uca.edu.ar/handle/123456789/9005
Resumen : Abstract: The protection and restoration of water-related ecosystems is one of the goals to be achieved by the United Nations’ 2030 Agenda for Sustainable Development. In this framework and requested by government Argentine institutions concerned with water, biodiversity and territorial management, this study analyzes the evolution of the flooded area within the Dulce River wetlands and Mar Chiquita Lake Nature Reserve (centered around 30.6°S, 62.6°W, 70m above sea level) in Argentina since 2003, when the historical maximum extent was reached, until 2017. The Modified Normalized Difference Water Index (MNDWI) was calculated on atmospherically corrected NASA Landsat 5 Thematic Mapper (L5-TM) and Landsat 8 Operational Land Imager (L8-OLI) reflectance data over two-scene cloudless-sky mosaics to cover the whole Reserve. Mixed-water pixels constituted an important fraction of the total-water covered area, particularly during years of minimum water level in Mar Chiquita Lake. So, MNDWI values were analyzed along transects crossing two stable regional water bodies to determine precise thresholds for detection of non-water (MNDWI < −0.15 for L5-TM, MNDWI < −0.35 for L8-OLI), mixedwater (−0.15 < MNDWI < 0.4 for L5-TM, −0.35 < MNDWI < 0.5 for L8-OLI) and open-water (0.4 < MNDWI for L5-TM, 0.5 < MNDWI for L8-OLI) pixels. A higher spatial resolution image, SPOT5-HGR2, was used to validate the classification method. A confusion matrix was built which resulted in an overall accuracy of 99.2% and a Kappa coefficient of 0.98. In-situ Geo-referenced photographic registers were also taken simultaneously to a Landsat 8 overpass to confirm the classification thresholds. The analysis of simulated MNDWI response, by using the assumption of the linear mixture model, showed that mixed pixels should present from 9% to 76% of detectable open-water area. Maximum total flooded area extensions of about 3600 km2 by 2003–2005 and a minimum one of 2050 km2 by the end of 2011 were established, followed by a recent trend to the recovering with a total flooded area of about 3400 km2 in the period 2015–2017. Open-water covered area follows closely the behavior of in-situ water level measurements of Mar Chiquita Lake, showing a maximum in year 2003 and a minimum towards the end of 2013, in a significant linear relation from which a topographical slope of the terrain of about 0.012% is inferred that agrees with previous bathymetric studies. Results show the powerful complement between a reliable water satellite monitoring tool and locally-measured parameters in so dynamic wetland regions.
Cobertura Espacial: Córdoba (Argentina: provincia)
URI : https://repositorio.uca.edu.ar/handle/123456789/9005
ISSN : 2352-9385
Disciplina: INGENIERIA AMBIENTAL
DOI: 10.1016/j.rsase.2019.04.003
Derechos: Acceso abierto. 2 años de embargo
Appears in Collections:Artículos

Files in This Item:
File Description SizeFormat
flooded-area-satellite-monitoring.pdf4,27 MBAdobe PDFThumbnail
View/Open
Show full item record

Page view(s)

175
checked on Mar 28, 2024

Download(s)

220
checked on Mar 28, 2024

Google ScholarTM

Check


Altmetric


This item is licensed under a Creative Commons License Creative Commons