Home |  English version |  Mappa |  Commenti |  Sondaggio |  Staff |  Contattaci Cerca nel sito  
Istituto sull'inquinamento atmosferico

Torna all'elenco Contributi in atti di convegno anno 2016

Contributo in atti di convegno

Tipo: Poster

Titolo: Assessment Of Atmospheric Correction Methods Of Sentinel-2 In Italian Lakes

Anno di pubblicazione: 2016

Formato: Cartaceo

Autori: Cazzaniga Ilaria (*), Bresciani Mariano (*), Giardino Claudia (*), Bassani Cristiana (**)

Affiliazioni autori: (*) CNR - IREA - Milano, (**) CNR - IIA

Autori CNR:

  • CRISTIANA BASSANI
  • MARIANO BRESCIANI
  • ILARIA CAZZANIGA
  • CLAUDIA GIARDINO

Lingua: inglese

Abstract: Accurate atmospheric correction products in inland water are required to retrieve water leaving reflectance and water quality parameters from Remote Sensing: a good estimate of water reflectance is mandatory to feed bio-optical models or empirical algorithms for the retrieval of concentrations of water constituents. The water leaving reflectance of lakes is also a new Essential Climate Variable and accurate estimates of this quantity are required by the Climate Change Initiative. The atmospheric correction algorithms used in this study and applied to Sentinel-2-MSI (S2) images are those recently proposed by EU projects (e.g. GLaSS, INFORM): a 6SV-based tool, ATCOR, SNAP-Sen2cor and ACOLITE. They were evaluated by using radiometric measurements gathered in few Italian lakes (i.e. Garda, Iseo, Maggiore, and Mantua fluvial lakes) with varying trophic conditions. A total of 9 images with corresponding 40 stations of in situ data were available. The images processing was accomplished as follow. The 6SV-based tool was used with standard parametrization, with fixed aerosol type for all the images and with Aerosol Optical Thickness (AOT) retrieved from AERONET sun-photometers measurements, or with AERONET aerosol microphysical properties for aerosol definition. ATCOR was run with 'Rural' aerosol type and visibility and water vapour fixed (with the same values used in the 6SV-based tool) or image-based estimated and variable within the scene. SNAP-Sen2cor was run with a 'Rural' aerosol type and with image-based estimated visibility. ACOLITE was used with SWIR bands combination, with bands ratio ? for aerosol contribution varying pixel-per-pixel. In few cases, Landsat-8-OLI (L8) images were acquired on the same day of S2, providing further data to be compared to S2. When compared to in situ data, the results on oligo-mesotrophic lakes, showed that the best Rrs estimate was performed by ACOLITE (mean ?2 was 0.04 and 0.02 respectively for Garda and Iseo lakes, taking into account the first nine S2 bands). For the more turbid productive waters of Mantua lakes, the best results in terms of Rrs estimate were obtained through 6SV-based tool and by ATCOR. Similar results were also obtained with L8 data. Then, by comparing S2 and L8 data acquired on the same days, the best match to field data was obtained for ACOLITE-S2 in Lake Iseo and with ATCOR-L8 in Mantua lakes. The results of this study are showing that for clearer waters (as those of Garda and Iseo lakes) the water-oriented processor ACOLITE is performing better than the land-oriented processors (such as ATCOR) that instead are performing well in more productive eutrophic waters. Nevertheless, further investigations are necessary to both augment the number of match-ups and to evaluate the uncertainties that is also associated to radiometric measurements gathered in the field.

Lingua abstract: inglese

Stato della pubblicazione: Published version

Parole chiave:

  • Atmospheric correction
  • chlorophyll-a
  • inland water
  • lakes
  • Correzione atmosferica
  • clorofilla-a
  • laghi
  • acque interne

Congresso nome: 1st Sentinel-2 Validation Team Meeting

Congresso luogo: ESRIN, Frascati, Roma

Congresso data: 28-29/11/2016

Congresso rilevanza: Internazionale

Congresso relazione: Contributo

Strutture CNR:

Moduli:

 
Torna indietro Richiedi modifiche Invia per email Stampa
Home Il CNR  |  I servizi News |   Eventi | Istituti |  Focus