European Facility For Airborne Research

European Facility For Airborne Research March 19, 2024, 04:23

JRA2 - Quality layers for airborne hyperspectral imagery and data products (HYQUAPRO)

JRA2 - Quality layers for airborne hyperspectral imagery and data products (HYQUAPRO)

HYQUAPRO sought to develop, implement and test quality indicators and quality layers for airborne hyperspectral imagery, and to develop higher performing water and soil algorithms as demonstrators for end-to-end processing chains with harmonised quality measures.

The concept of Uncertainty Propagation Analysis (UPA) combined with Monte Carlo stochastic simulation has been applied to airborne hyperspectral imagery to explore how uncertainty of input parameters propagates through the processing chain.

Within the consortium of 9 European data providers for airborne hyperspectral imagery (VITO/UZH, DLR, INTA, PML, ISBE, TAU and FUB, with ONERA associated), quality indicators were identified and selected for implementation in the processing chains. Also the data description that accompanies the hyperspectral data (metadata) has been harmonised across all EUFAR providers. The main step was the development of data quality indicators. Therefore generic quality indicators and quality layers for airborne hyperspectral images (based on sensor and scene characteristics) were developed. Sensor characteristics, sensor calibration, data characterisation, sensor performance during acquisition, external conditions during acquisition, quality of auxiliary data used for the processing were translated into generic quality indicators and quality layers. Starting from the generic quality layers, these were adjusted (“personalised”) for the different processing facilities involved since different sensors, software, auxiliary data and processing methods are used at the different processing facilities. After adjustment of the quality layers, the layers were integrated in the respective processing facilities.

A literature review of water quality algorithms, from simple to complex approaches was undertaken and preliminary results were presented at the EWG meeting on “Water Applications” during RP1. PML continued with the algorithm review following advice on variables to consider from the EWG. PML investigated the development of an improved version of an Inherent Optical Properties model for use in inland and coastal waters.

For the integration of the PML higher performing water quality algorithms into an existing PAF, three algorithms were provided by PML to VITO:

      • Algorithm of Gitelson et al. (2007) for CHL-a retrieval
      • SeaWiFS OC4v6 algorithm of O’Reilly et al., 2000 for CHL-a retrieval
      • IOP model of Smyth et al. (2006) for IOP retrieval.

The integration of these algorithms in the PAF was tested with the Lake Balaton 2010 data set (NERC-ARSF AISA Eagle, in-situ and sun photometer data) provided by PML and the University of Stirling.

The validation of water quality retrieval algorithms implemented by the VITO PAF processing chain was implemented at the stages of atmospheric correction and retrieval of water quality indicators, such as chlorophyll-a concentration, water IOPs and TSM concentration.

During RP4 (April 2013 – Sept 2013), the PML inherent optical property model was updated to include optical water types with considerable improvements in retrieval of the absorption (in terms of RMSE). The backscatter was also improved but to a lesser extent in RMSE. The algorithm has since been optimised resulting in a significant improvement in processing time (a factor of 23 faster).

HYQUAPRO developments allow provision of quality indicators/quality layers (and data descriptors/metadata) with the hyperspectral imagery to their users. In addition, validated water quality products (Chl-a, IOP) are now available to the users through the VITO processing facility. Partner GFZ undertook the development of higher performing soil algorithms under the double commitment of using methodologies where automation is possible, and offering multiple algorithms to the users. The focus was on offering both analytical and empirical algorithms for the determination of the following key soil products: clay, iron, carbonate, soil organic carbon, and soil moisture maps.

The HYperspectral SOil MApper (HYSOMA) toolbox was developed and validated with 18 image datasets and allows to produce 11 soil products associated with different methods for soil moisture content, soil organic carbon content, and soil minerals content (iron oxides, clay, carbonates) for every input image file, plus 1 soil quality layer file, and 4 mask files. For the integration of the GFZ higher performing soil algorithms into an existing PAF an automatic version of HYSOMA was developed under the name HYSOMA_AUTO. HYSOMA_AUTO runs without interface under the IDL command prompt and was integrated in the automated DLR processing chain. The validation of the HYSOMA products based on various in-situ validation data sets showed correlations from R2 of 0.52 (clay) up to >0.9 (soil moisture) which both validate the HYSOMA software and provide science validation for the soil algorithms.

The HYSOMA software was adapted to be included in the EUFAR Toolbox. For this, a public release version was developed and the HYSOMA website (www.gfz-potsdam.de/hysoma) was released on 29 June 2012, to which the EUFAR Toolbox is directly linked. After registration and accepting the license, HYSOMA is freely made available for download for non-commercial purposes through both the N6SP EUFAR Toolbox and the HYSOMA website. Plug-ins for linux/mac/windows can be found on this website. The software is IDL based and distributed for free under the IDL-virtual machine, so that it is easy to use for non-expert users. At the end of the project in 2013, more than 60 users from all over the world had downloaded a plug-in of the HYSOMA software interface, demonstrating the interest of the airborne community for this software.

The PML IOP algorithm and the other water quality algorithms are freely available through the EUFAR toolbox.

For more information, contact the activity leader - Ils Reusen (ils.reusen@vito.be).

Related publications:

Bachmann, M., Makarau, A., Segl, K., Richter, R. (2015): Estimating the influence of spectral and radiometric calibration uncertainties on EnMAP data products - examples for ground reflectance retrieval and vegetation indices. Remote Sensing, 7.

Bachmann, M., Rogge, D., Malec, S., Holzwarth, S., Makarau, A., Richter, R. (2015) Estimating the uncertainty in ground reflectances resulting from radiometric and spectral calibration. EARSeL SIG-IS, 14.-16.04.2015, Luxembourg.

Nitin Bhatia, Valentyn A. Tolpekin, Ils Reusen, Sindy Sterckx, Jan Biesemans, and Alfred Stein (2015). Sensitivity of Reflectance to Water Vapor and Aerosol Optical Thickness. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, VOL. 8, NO. 6, June 2015, pp. 3199-3208. 10.1109/JSTARS.2015.2425954

Johan Beekhuizen, Gerard B. M. Heuvelink, Jan Biesemans, and Ils Reusen (2011). Effect of DEM Uncertainty on the Positional Accuracy of Airborne Imagery. IEEE Transactions on Geoscience and Remote Sensing, VOL. 49, NO. 5, May 2011, pp. 1567-1577. 10.1109/TGRS.2010.2083672  


Page last modified on Sept. 12, 2017, 10:46
Back to top
Copyright © 2024 EUFAR All rights reserved.