Elsevier

Remote Sensing of Environment

Volume 175, 15 March 2016, Pages 183-195
Remote Sensing of Environment

Contrasting snow and ice albedos derived from MODIS, Landsat ETM+ and airborne data from Langjökull, Iceland

https://doi.org/10.1016/j.rse.2015.12.051Get rights and content

Highlights

  • 5 m resolution airborne dataset compared to Landsat ETM+ and MODIS datasets

  • Snow and ice albedo measurements over Langjökull, Iceland compared

  • Different albedo retrieval methods produce inconsistent albedo measurements.

  • Contrasts in albedo measurements associated with specific facies types

Abstract

Surface albedo is a key parameter in the energy balance of glaciers and ice sheets because it controls the shortwave radiation budget, which is often the dominant term of a glacier's surface energy balance. Monitoring surface albedo is a key application of remote sensing and achieving consistency between instruments is crucial to accurate assessment of changing albedo. Here we take advantage of a high resolution (5 m) airborne multispectral dataset that was collected over Langjökull, Iceland in 2007, and compare it with near contemporaneous ETM+ and MODIS imagery. All three radiance datasets are converted to reflectance by applying commonly used atmospheric correction schemes: 6S and FLAASH. These are used to derive broadband albedos. We first assess the similarity of albedo values produced by different atmospheric correction schemes for the same instrument, then contrast results from different instruments. In this way we are able to evaluate the consistency of the available atmospheric correction algorithms and to consider the impacts of different spatial resolutions. We observe that FLAASH leads to the derivation of surface albedos greater than when 6S is used. Albedo is shown to be highly variable at small spatial scales. This leads to consistent differences associated with specific facies types between different resolution instruments, in part attributable to different surface bi-directional reflectance distribution functions. Uncertainties, however, still exist in this analysis as no correction for variable bi-directional reflectance distribution functions could be implemented for the ETM+ and airborne datasets.

Introduction

A key concern associated with rising high northern latitude temperatures is the melting of terrestrial ice bodies leading to the rise of global sea levels (Dowdeswell et al., 1997, Hagen et al., 2003, Meier et al., 2007, Radić and Hock, 2011, Wolken et al., 2009). Arctic and sub-Arctic ice masses are particularly sensitive to climate change as temperatures there are rising at about twice the global average (Graversen, Mauritsen, Tjernström, Källén, & Svensson, 2008). Predicting the response of terrestrial ice bodies to high northern latitude climate change requires accurate calculation of surface melt rates and thus precise assessment of the ice surface energy-balance (Aas et al., 2015). In many systems, energy balance studies have shown that net shortwave radiation is often the dominant contributor of available energy for melting glacier snow and ice (Arendt, 1999). The amount of energy available to glacier surfaces from shortwave fluxes is controlled by the surface albedo, i.e. its reflectivity. Accurate measurement and parameterisation of surface albedo is therefore a key component in calibrating/validating energy balance models designed to estimate past, current and future glacier melt. This is particularly important in Arctic and sub-Arctic regions because a key reason why high northern latitude temperatures are rising so rapidly is due to albedo feedbacks (Serreze, Holland, & Stroeve, 2007).

Glacier surface reflectance can be measured using either in situ ground based methods or remote sensing techniques (Cutler and Munro, 1996, Reijmer et al., 1999). Satellite and airborne remote sensing allows both large spatial coverage and regular temporal sampling, and limits the cost and risk associated with repeat field measurements (Aniya et al., 1996, Box et al., 2012, Boyd, 2009, Paul et al., 2004). The various instruments on board different remote sensing platforms use a range of spectral bands which require geometric and atmospheric correction to convert radiance to reflectance, and Narrow-To-Broadband (NTB) transformation to produce average surface reflectance. The suitability of individual sensors for energy balance studies depends, therefore, on the spectral and spatial resolution of individual sensors, the reliability of geometric/atmospheric correction techniques, and the precision of the NTB transformation (Chander et al., 2009, Greuell and Oerlemans, 2004, Rees, 2006, Vermote et al., 2002).

The albedo of glacier snow and ice is highly spatially and temporally variable. It depends on a range of factors including solar incidence angle, cloud cover, surface topography, snow grain size and geometry, impurities in the snow and ice, and water content (Arnold et al., 2006, Dumont et al., 2012, Warren and Wiscombe, 1980, Wiscombe and Warren, 1980). Albedo varies over spatial scales on the order of metres (Arnold & Rees, 2003) and it evolves temporally as snow metamorphoses and melts or as new snow falls (Brock, Willis, Sharp, & Arnold, 2000). Surface albedo and its association with different snow and ice facies has been monitored throughout the year in a range of climatic settings (Box et al., 2012, Klok and Oerlemans, 2002, Tedesco et al., 2011). Fresh snow can have an albedo of over 0.9 whereas ice can range between ~ 0.4 and ~ 0.1 depending on the debris content (Cuffey & Paterson, 2010).

Many studies have attempted to compare satellite derived albedos with ground measurements and to improve the algorithms used to derive surface albedo from satellite data (Greuell et al., 2002, Hall et al., 1989, Hendriks and Pellikka, 2004, Knap and Reijmer, 1998, Liang et al., 2005, Reijmer et al., 1999, Stroeve et al., 2005). Such studies use mainly ground point measurements and are therefore limited by their lack of spatial and temporal resolution. Regarding spatial resolution, point measurements have sometimes been used to validate measurements for pixels with areas up to 1 km2. Despite their frequent use, point measurement validations rely on relatively homogeneous surface reflectance characteristics in the surrounding areas. This is frequently not the case, especially for valley glaciers and the outlet glaciers from ice caps and ice sheets.

Greuell et al. (2002) sought to improve on ground point measurements using helicopter-based albedo readings. However, the spatial coverage of this study was still rather limited, equating to a few thousand data points. By contrast, in the present study, we use 1.32 × 107 measurements derived from a high resolution Airborne Thematic Mapper (ATM) dataset to validate both Landsat 7 Enhanced Thematic Mapper Plus (ETM+) and MODerate Resolution Imaging Spectroradiometer (MODIS) albedo data. The primary aims of our study are to assess the spatial pattern of surface albedo across an ice mass and to evaluate the consistency of different scientific products derived from various remote sensing instruments. The study is applied to Langjökull, a typical Icelandic ice mass, which exhibits a large range of albedos both spatially across the ice cap and temporally through the year.

Section snippets

Study site

Langjökull (64.7°N, 20.4°W) is Iceland's second largest ice cap, with an area of ~ 925 km2 (Fig. 1). The ice cap elevation ranges from 450 to 1450 m above sea level with an average height of 900 m (Pope, Willis, Rees, Arnold, & Palsson, 2013). Langjökull is surrounded by basalt lava fields, sandur plains and proglacial lakes and while two major rivers drain some meltwater from the ice cap, a significant proportion drains directly into the substrate to feed groundwater aquifers (Guðmundsson,

Results

Results are divided into two parts. First, for both the ATM and ETM+ datasets, the effects of the different processing techniques (i.e. FLAASH and 6S) are analysed. Second, differences between the albedo estimates generated by the different instruments (ATM, ETM+ and MODIS) are compared. The spatial extent of all the datasets is the same, as each dataset has been masked to the full extent of available albedo values across all datasets.

Discussion

The discussion is divided into three parts. First, we outline the possible reasons for the differences between the ATM6S, ETM6S and MCD43 datasets. Second, we evaluate the MCD43 product against the ATM6S and ETM6S datasets. Last, we assess the implications of the differences between the datasets for energy balance modelling and melt estimates.

Conclusions

This study has explored the ability of different resolution instruments and different retrieval methods to measure the surface albedo across Langjökull. Different retrieval methods for the same instrument have been shown to produce inconsistent surface albedo measurements. These differences are the result of contrasts between different atmospheric correction models which were applied. Correction of both ATM and ETM+ datasets using FLAASH produced mean albedos greater than those generated by 6S.

Acknowledgements

This work was supported by the UK NERC ARSF — Project IPY07-08. E. Pope was supported by the NERC Arctic Research Programme under project NE/K00008Xs/1. A. Pope was supported by the National Science Foundation Graduate Research Fellowship Programme under Grant No. DGE-1038596 and by Trinity College, Cambridge. E. Miles was supported by a Gates Cambridge Scholarship and by Trinity College, Cambridge. Fieldwork associated with the ATM flights was supported by grants from the University of

References (72)

  • A. Marks et al.

    Characterisation of the HDRF (as a proxy for BRDF) of snow surfaces at Dome C, Antarctica, for the inter-calibration and inter-comparison of satellite optical data

    Remote Sensing of Environment

    (2015)
  • C.B. Schaaf et al.

    First operational BRDF, albedo nadir reflectance products from MODIS

    Remote Sensing of Environment

    (2002)
  • G. Schaepman-Strub et al.

    Reflectance quantities in optical remote sensing—Definitions and case studies

    Remote Sensing of Environment

    (2006)
  • J. Stroeve et al.

    Accuracy assessment of the MODIS 16-day albedo product for snow: Comparisons with Greenland in situ measurements

    Remote Sensing of Environment

    (2005)
  • J. Stroeve et al.

    Re-evaluation of MODIS MCD43 Greenland albedo accuracy and trends

    Remote Sensing of Environment

    (2013)
  • E.F. Vermote et al.

    Atmospheric correction of MODIS data in the visible to middle infrared: First results

    Remote Sensing of Environment

    (2002)
  • K.S. Aas et al.

    A comparison between simulated and observed surface energy balance at the Svalbard archipelago

    Journal of Applied Meteorology and Climatology

    (2015)
  • M. Aniya et al.

    The use of satellite and airborne imagery to inventory outlet glaciers of the Southern Patagonia Icefield, South America

    Photogrammetric Engineering and Remote Sensing

    (1996)
  • A. Arendt

    Approaches to modelling the surface albedo of a high Arctic glacier

    Geografiska Annaler Series a-Physical Geography

    (1999)
  • N.S. Arnold et al.

    Self-similarity in glacier surface characteristics

    Journal of Glaciology

    (2003)
  • N.S. Arnold et al.

    Topographic controls on the surface energy balance of a high Arctic valley glacier

    Journal of Geophysical Research — Earth Surface

    (2006)
  • H. Björnsson et al.

    Icelandic glaciers

    Jökull

    (2008)
  • J.E. Box et al.

    Greenland ice sheet albedo feedback: Thermodynamics and atmospheric drivers

    The Cryosphere

    (2012)
  • D.S. Boyd

    Remote sensing in physical geography: A twenty-first-century perspective

    Progress in Physical Geography

    (2009)
  • B.W. Brock et al.

    Modelling seasonal and spatial variations in the surface energy balance of Haut Glacier d'Arolla, Switzerland

    Annals of Glaciology

    (2000)
  • K.M. Cuffey et al.

    The physics of glaciers

    (2010)
  • P.M. Cutler et al.

    Visible and near-infrared reflectivity during the ablation period on Peyto Glacier, Alberta, Canada

    Journal of Glaciology

    (1996)
  • M. Dumont et al.

    Variational assimilation of albedo in a snowpack model and reconstruction of the spatial mass-balance distribution of an alpine glacier

    Journal of Glaciology

    (2012)
  • R.G. Graversen et al.

    Vertical structure of recent Arctic warming

    Nature

    (2008)
  • S. Guðmundsson et al.

    Comparison of energy balance and degree-day model of summer ablation on the Langjökull ice cap, SW-Iceland

    Jökull

    (2009)
  • J.O. Hagen et al.

    On the net mass balance of the glaciers and ice caps in Svalbard, Norwegian Arctic

    Arctic, Antarctic, and Alpine Research

    (2003)
  • J. Hendriks et al.

    Estimation of reflectance from a glacier surface by comparing spectrometer measurements with satellite-derived reflectances

    Journal of Glaciology

    (2004)
  • R.A. Hill et al.

    Mapping tree species in temperate deciduous woodland using time-series multi-spectral data

    Applied Vegetation Science

    (2010)
  • R. Hock et al.

    A distributed surface energy-balance model for complex topography and its application to Storglaciären, Sweden

    Journal of Glaciology

    (2005)
  • E.J. Klok et al.

    Model study of the spatial distribution of the energy and mass balance of Morteratschgletscher, Switzerland

    Journal of Glaciology

    (2002)
  • W.H. Knap et al.

    Comparison of Landsat TM-derived and ground-based albedos of Haut Glacier d'Arolla, Switzerland

    International Journal of Remote Sensing

    (1999)
  • Cited by (23)

    View all citing articles on Scopus
    View full text