PosterPDF Available

Validation of the Operational Land Imager orange contra-band retrieval for inland water quality applications

Authors:
Alexandre Castagna at Ghent University
Alexandre Castagna
Stefan G. H. Simis at Plymouth Marine Laboratory
Stefan G. H. Simis
Heidi M Dierssen at University of Connecticut
Heidi M Dierssen
Quinten Vanhellemont at Royal Belgian Institute of Natural Sciences
Quinten Vanhellemont
Show all 6 authorsHide
Download file PDF
Read file
Download citation

Abstract

This poster presents the results of a validation study on the calculated orange contra-band (613 nm) for OLI/Landsat 8 against the orange band (620 nm) of OLCI/Sentinel-3. The approach forces atmospheric compensation (AC) bias to be equal for each sensor and scene, allowing to compare the retrieved water leaving signal directly. We find that OLI orange contra-band has the expected relation to OLCI/Sentinel-3 as observed with in situ data for all studied lakes, except one. This further evidence that the contra-band algorithm has a solid analytical foundation and is robust to sensor noise and AC impact on the OLI input bands. Final peer-review study available here: https://doi.org/10.3390/rs12040637
Content uploaded by Alexandre Castagna
Author content
All content in this area was uploaded by Alexandre Castagna on May 20, 2019
Content may be subject to copyright.
Expectation from in situ data: slope between 0.85
and 1, depending on the presence of phycocyanin.
Validation of the Operational Land Imager orange contra-band
retrieval for inland water quality applications
Acknowledgments
Matchups with OLCI/Sentinel-3
Alexandre Castagna, Stefan Simis, Heidi Dierssen, Quinten Vanhellemont, Koen Sabbe, Wim Vyverman
Partners
Motivation
A m e t h o d i s n o w a v a i l a b l e t o extract
a d d i t i o n a l s p e c t r a l i n f o r m a t i o n f r o m
overlapping wavebands (Castagna et al.,
2018). When applied to OLI/Landsat 8, an
o r a n g e c o n t r a - b a n d ( 6 1 3 n m ) c a n b e
retrieved from the Green, Red and Pan
b a n d s . T h i s c o n t r a - b a n d c o v e r s t h e
phycocyanin absorption peak and can be
used for cyanobacteria detection at high
spatial resolution. Here we further evaluate
its performance with comparison against
OLCI/Sentinel-3 orange band (620 nm),
taken as a calibrated reference. The goal is
to validate the absolute magnitude of the
orange contra-band.
C a s t a g n a e t a l . , 2 0 1 8 , E x t e n d i n g t h e
O p e r a t i o n a l L a n d I m a g e r / L a n d s a t 8 f o r
freshwater research: retrieval of an orange
band from PAN and MS bands. Ocean Optics
XXIV, Dubrovinik, Croatia.
Example application
Sediment
Cyanobacteria
Floating vegetation
Validation approach
Embalse Cerron Grande, SV, 2018-09-06
To compare the absolute reflectance magnitude
be tw ee n OL I and OLCI it i s n ec es s ar y tha t
at m os p her ic e ff e ct s a re rem o ved and any
residual atmospheric bias be equal for both
sensors. Compensation for atmosphere effects is
challenging over turbid inland waters in the
absence of SWIR bands, therefore OLI processed
with ACOLITE was taken as reference in the
multispectral bands other than the orange. OLCI
imagery was first partially compensated for
Rayleigh effects only with SeaDAS and then
subtracted by the median difference to OLI
multispectral bands. The OLCI orange band (620
nm) was corrected with the interpolated median
difference between Green and Red bands. Spatial
aggregation and bandshifting was performed
before the correction.
alexandre.castagna@ugent.be
80 matchups between OLI and OLCI were
processed, covering 10 lakes between 2016
and 2019.
Check out imagery
and more at:
... While the results of that [5] study are promising, their approach relied solely on an in situ dataset and simulated OLI data. In a second study, Castagna et al. [6] used satellite data to validate the OLI's virtual band with same-date OLCI data. However, their study did not show a comparison between the estimated PC maps derived from the new virtual orange band and PC maps using the OLCI's actual orange band. ...
Landsat 8 Virtual Orange Band for Mapping Cyanobacterial Blooms
Article
Full-text available
  • Mar 2020
The Landsat 8 Operational Land Imager (OLI) has a panchromatic band (503–676 nm) that can be used to derive a novel virtual orange band (590–635 nm) by using the multispectral green band and red band components. The orange band is useful for the accurate detection and quantification of phycocyanin (PC), an accessory pigment in toxin-producing cyanobacterial blooms, because of the specific light absorption characteristics of PC around 600–625 nm. In this study, we compared the Landsat 8 OLI’s and Sentinel-3 Ocean and Land Color Instrument’s (OLCI) derived orange band reflectance and PC products corresponding to a same-date overpass during a severe cyanobacterial bloom in Lake Erie, USA. The goal was to determine if the OLI’s virtual orange band can produce results equivalent to the OLCI’s actual orange band. Band-by-band match-ups used the OLI’s top-of-atmosphere (TOA) reflectance versus TOA reflectance from the OLCI, and surface reflectance (SR) from the OLI versus SR from the OLCI. A significant correlation was observed between the OLI’s and OLCI’s derived orange band TOA reflectance (R2 = 0.86; p < 0.001; NRMSE = 9.01%) and orange band SR (R2 = 0.93; p < 0.001; NRMSE = 20.23%). The PC map produced using the best-fit empirical models from both sensors showed similar PC spatial patterns and concentration levels in the western basin of Lake Erie. The results from this research are particularly important for the study of smaller inland waterbodies with the 30 m resolution of the OLI, which cannot be studied with the 300 m resolution of OLCI data, and for analyzing historical bloom events before the launch of the OLCI. Although more analysis and validation need to be conducted, this study opens up Landsat 8’s applicability in research on cyanobacterial harmful algal blooms (cyanoHABs).
View
ResearchGate has not been able to resolve any references for this publication.