Elsevier

Remote Sensing of Environment

Volume 115, Issue 8, 15 August 2011, Pages 2126-2140
Remote Sensing of Environment

Accuracy assessment of land surface temperature retrievals from MSG2-SEVIRI data

https://doi.org/10.1016/j.rse.2011.04.017Get rights and content

Abstract

The accuracy of the Land Surface Temperature (LST) product generated operationally by the EUMETSAT Land Surface Analysis Satellite Applications Facility (LSA SAF) from the data registered by the Spinning Enhanced Visible and Infrared Imager (SEVIRI) on board the geostationary METEOSAT Second Generation 2 (MSG2, Meteosat 9) satellite was assessed on two test sites in Eastern Spain: a homogeneous, fully vegetated rice field and a high-plain, homogeneous area of shrubland. The LSA SAF LSTs were compared with ground LST measurements in the conventional temperature-based (T-based) method. We also validated the LSA SAF LST product by using an alternative radiance-based (R-based) method, with ground LSTs calculated from MSG-SEVIRI channel 9 brightness temperatures (at 10.8 μm) through radiative transfer simulations using atmospheric temperature and water vapor profiles together with surface emissivity data. Two lakes were also used for validation with the R-based method. Although the LSA SAF LST algorithm works mostly within the uncertainty expectation of ± 2 K, both validation methods showed significant biases for the LSA SAF LST product, up to 1.5 K in some cases. These biases, with the LSA SAF LST product overestimating reference values, were also observed in previous studies. Nevertheless, the present work points out that the biases are related to the land surface emissivities used in the operational generation of the product. The use of more appropriate emissivity values for the test sites in the LSA SAF LST algorithm led to better results by decreasing the biases by 0.7 K for the shrubland validation site. Furthermore, we proposed and checked an alternative algorithm: a quadratic split-window equation, based on a physical split-window model that has been widely proved for other sensors, with angular-dependent coefficients suitable for the MSG coverage area. The T-based validation results for this algorithm showed LST uncertainties (robust root-mean-squared-errors) from 0.2 K to 0.5 K lower than for the LSA SAF LST algorithm after the emissivity replacement. Nevertheless, the proposed algorithm accuracies were significantly better than those obtained for the current LSA SAF LST product, with an average accuracy difference of 0.6 K.

Research highlights

► The accuracy of the EUMETSAT LSA SAF LST product was assessed (T- & R-based methods). ► Both validation methods showed significant biases for the LSA SAF LST product. ► The biases can be due to the land surface emissivities used as input data. ► An alternative LST algorithm is proposed and checked, showing better accuracies.

Introduction

Land Surface Temperature (LST) is defined as the temperature of the interface between the Earth's surface and its atmosphere and thus it is a critical variable to understand land–atmosphere interactions and a key parameter in a wide range of environmental applications, such as meteorological, climatological and hydrological studies, which involve energy and water fluxes (Kustas and Norman, 1999, Sánchez et al., 2009), irrigation need estimates (Brasa-Ramos et al., 1998), forest fire risk forecasting, detection and monitoring (Briess et al., 2003), air temperature and humidity determination (Florio et al., 2004, Vogt et al., 1997), leaf wetness evaluation (Deshpande et al., 1995) and desertification, deforestation and climate change monitoring (Allen et al., 1994, Lambin and Ehrlich, 1997). Additionally, a frequent and accurate determination of LST could help to improve the forecasting of natural hazards and the analysis of thermal diurnal cycles.

Thermal infrared (TIR) remote sensing is used to obtain LST measurements at different spatial and temporal resolutions. Examples are the LST products generated from the Advanced Along-Track Scanning Radiometer (AATSR) (Llewellyn-Jones et al., 2001) onboard the European ENVISAT, as well as the Moderate Resolution Imaging Spectroradiometer (MODIS) (Wan, 2008) and the Advanced Spaceborne Thermal Emission and Reflection radiometer (ASTER) (Gillespie et al., 1998) on EOS platforms. The Spinning Enhanced Visible and Infrared Imager (SEVIRI) on board the geostationary METEOSAT Second Generation (MSG) platform allows a semi-global coverage with a higher temporal resolution (one image every 15 min) and thus offers the possibility to provide more frequent LST determinations and to study the diurnal LST cycle. An operational MSG-SEVIRI LST product is generated by the Land Surface Analysis of the Satellite Application Facility (LSA SAF) at the European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT).

The aim of this paper is to evaluate the accuracy of LSTs retrieved with predefined MSG algorithms before making use of the high temporal resolution of MSG-SEVIRI data. The paper first describes the operational LSA SAF procedure to generate MSG LST images. Then, the paper proposes an alternative algorithm with a physical theoretical background that corrects atmospheric effects by using angular-dependent coefficients. Both the LSA SAF and the proposed algorithms use land surface emissivity (LSE) as input data, and previous accurate LSE maps are thus required since emissivity uncertainties can involve important errors in terms of LST. The sensitivity of the proposed algorithm to different uncertainty sources is analyzed in Section 3. The accuracy of the proposed algorithm and the LSA SAF product are then checked in Section 4 by using concurrent in situ data at two dedicated experimental sites in Eastern Spain, and also by means of a new radiance-based validation method (Wan & Li, 2008) that does not require ground LST measurements and thus avoids possible problems of spatial scale dissimilarity between ground and satellite data. Finally, the main conclusions are summarized in Section 5.

Section snippets

The operational MSG-SEVIRI LST algorithm

The LSA SAF LST product is generated by using a Generalized Split-Window (GSW) algorithm (Madeira, 2002, Trigo et al., 2008a) that follows the formulation first proposed by Wan and Dozier (1996) for data from the Advanced Very High Resolution Radiometer (AVHRR) on the National Oceanic and Atmospheric Administration (NOAA) platforms and EOS-MODIS. The LST is a function of the cloud-free top-of-atmosphere brightness temperatures measured by MSG-SEVIRI channels 9 and 10, centered at 10.8 μm and 12 

Sensitivity analysis

The uncertainty of the algorithm proposed in Eq. (4) was evaluated with a sensitivity analysis, similarly as in Freitas et al. (2010). LST uncertainty was estimated as the combination of the model error, δ(LST)M, and the LST uncertainties associated with the input uncertainties, δ(LST)P:δ(LST)=δ(LST)M2+δ(LST)P21/2where we assumed that both errors are independent. δ(LST)M is defined as:δ(LST)M=σAC2+(1ε)σα2+Δεσβ21/2and δ(LST)P follows:δ(LST)P=iLSTxiδxi21/2σAC being the fitting error of the

Validation

Trigo et al., 2008a, Kabsch et al., 2008 quantified LSA SAF LST accuracy by using ground data collected from a validation station in a Quercus woodland plain near Evora, Portugal (38.54°N, 8.00°W). Accuracies from ± 1.0 K to ± 4.9 K were obtained depending on the season and the selection of daytime/night-time data, and the existence of significant biases between satellite-retrieved and in situ LSTs was pointed out. These inaccuracies could be a consequence of the mentioned spatial-scale

Conclusions

The operational LST product generated every 15 min by the EUMETSAT LSA SAF from MSG-SEVIRI data was first tested at two dedicated, partly and fully vegetated ground-truth sites in Eastern Spain, and LST uncertainties larger than ± 1.5 K were observed as a consequence of the existence of significant biases. Eq. (4), which is a quadratic split-window equation with angular-dependent coefficients, was proposed as an alternative algorithm. This equation follows the theoretical split-window model first

Acknowledgements

The research described in this paper was supported by the Spanish Ministerio de Ciencia e Innovación (projects CGL2007-65774/CLI, CGL2008-04550, CGL2010-16364, and CONSOLIDER-INGENIO 2010 CSD2007-00067; and Dr. Niclòs' Research Contract funded by the “Subprograma Ramón y Cajal" (MICINN-RYC)), the Conselleria d'Educació of the Generalitat Valenciana (PROMETEO/2009/086), and the European Union (Project CIRCE No. 036961). Instituto Universitario CEAM-UMH is supported by the Generalitat Valenciana

References (54)

  • Z. Wan

    New refinements and validation of the MODIS Land-Surface Temperature/Emissivity products

    Remote Sensing of Environment

    (2008)
  • M.R. Allen et al.

    Global change detection

    Nature

    (1994)
  • A. Berk et al.

    MODTRAN 4 User's Manual

    (1999)
  • K. Briess et al.

    Fire recognition potential of the bi-spectral infrared detection (BIRD) satellite

    International Journal of Remote Sensing

    (2003)
  • V. Caselles et al.

    Thermal band selection for the PRISM instrument. 1. Analysis of emissivity-temperature separation algorithms

    Journal of Geophysical Research

    (1997)
  • CEC (Commission of the European Communities)

    CORINE Land Cover Technique Guide

    (1995)
  • F. Chevallier et al.

    TIGR-like atmospheric profile database for accurate radiative flux computation

    Quarterly Journal of the Royal Meteorological Society

    (2000)
  • C. Coll et al.

    A split-window algorithm for land surface temperature from Advanced Very High Resolution Radiometer data: Validation and algorithm comparison

    Journal of Geophysical Research

    (1997)
  • C. Coll et al.

    Temperature-based and radiance-based validations of the V5 MODIS land surface temperature product

    Journal of Geophysical Research

    (2009)
  • Coll, C., Valor, E., Galve, J. M., Mira, M., Bisquert, M., García-Santos, V., et al. (in press). Long-term accuracy...
  • R.Y. Deshpande et al.

    Estimating leaf wetness in dry bean canopies as a prerequisite to evaluating white mold disease

    Agronomy Journal

    (1995)
  • E.N. Florio et al.

    Integration AVHRR satellite data and NOAA ground observations to predict surface air temperature: a statistical approach

    International Journal of Remote Sensing

    (2004)
  • S.C. Freitas et al.

    Quantifying the uncertainty of land surface temperature retrievals from SEVIRI/Meteosat

    IEEE Transactions on Geoscience and Remote Sensing

    (2010)
  • J.M. Galve et al.

    An atmospheric radiosounding database for generating land surface temperature algorithms

    IEEE Transactions on Geoscience and Remote Sensing

    (2008)
  • B.C. Gao et al.

    The MODIS Near-IR Water Vapor Algorithm, Product ID: MOD05—Total Precipitable Water Algorithm Technical Background Document

    (1998)
  • F.J. García-Haro et al.

    Variable multiple endmember spectral mixture analysis (VMESMA)

    International Journal of Remote Sensing

    (2005)
  • A.R. Gillespie et al.

    Temperature and emissivity separation from Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) images

    IEEE Transactions on Geoscience and Remote Sensing

    (1998)
  • Cited by (0)

    View full text