Elsevier

Geothermics

Volume 72, March 2018, Pages 156-162
Geothermics

Australian mean land-surface temperature

https://doi.org/10.1016/j.geothermics.2017.10.008Get rights and content

Highlights

  • A mean land-surface temperature map of the Australian continent was produced from 13 years of MODIS satellite imagery, for the period 2003–2015.

  • The results show good agreement with independent methods of estimating average land-surface temperature.

  • In comparison to previously used methods of estimating mean land-surface temperature, our new estimates are up to 12 °C warmer.

  • The map provides spatially continuous estimates of land-surface temperature that can be incorporated as the surface thermal boundary condition in geothermal studies.

Abstract

The mean land-surface temperature represents an important boundary condition for many geothermal studies. This boundary is particularly important to help constrain the models made to analyse resource systems, many of which are shallow in nature and observe relatively small thermal gradients. Consequently, a mean land-surface temperature map of the Australian continent has been produced from 13 years of MODIS satellite imagery, for the period 2003–2015. The map shows good agreement with independent methods of estimating mean land-surface temperature, including borehole surface-temperature extrapolation and long-term, near-surface ground measurements. In comparison to previously used methods of estimating mean land-surface temperature, our new estimates are up to 12 °C warmer. The MODIS-based method presented in this study provides spatially continuous estimates of land-surface temperature that can be incorporated as the surface thermal boundary condition in geothermal studies. The method is also able to provide a quantification of the uncertainties expected in the application of these estimates for the purposes of thermal modelling.

Introduction

The thermal budget of the crust forms an important element in our overall understanding of geological systems. Indeed, while the embedded thermal energy of the crust represents an energy resource of itself, crustal temperature is also an important constraint on geodynamic modelling, hydrogeochemical modelling, the accurate interpretation and inversion of geophysical data sets, and the formation and preservation of both petroleum and mineral systems (e.g. Beardsmore and Cull, 2001, Clauser, 2003, Davies, 1999, Magoon and Dow, 1994, Sandiford et al., 2002, Wyborn et al., 1994). In particular, the shallow thermal models used for the analysis of resource systems may only be interested in subsurface temperatures of up to 100–200 °C. Errors of a few degrees can have significant impacts on the interpretations drawn from such models (Peters and Nelson, 2012).

Conductive thermal models of the crust are subject to many sources of uncertainty. These include uncertainty in the geological structure, rock properties (thermal conductivity and heat production), and in the model boundary conditions (typically fixed basal heat flow and fixed surface temperature). The focus of this paper is on the specific choice of a fixed temperature to apply as the surface boundary condition. Previous studies have already highlighted the importance of this parameter, particularly on the sensitivity of results from shallow thermal models (c.f. Kohl et al., 2001). However, in the absence of detailed ground-temperature sampling, studies to-date have generally had to rely on the use of proxy data.

In the past, many researchers have simply adopted the average annual air temperature as the surface boundary constraint (e.g. Middleton, 1979, Chapman et al., 1984, Goutorbe et al., 2008, Danis et al., 2010). Others have leveraged the adiabatic lapse rate of air to estimate topography-dependent empirical corrections for known ground-surface temperatures (e.g. Kohl et al., 2001, Kohl et al., 2003). However, air temperature is only one of the variables influencing ground temperature, and its use has a proxy has long been recognised to introduce error. Howard and Sass (1964) compared ground surface temperature values, derived from borehole thermal gradient extrapolations, with mean annual air temperature values for 11 boreholes across the Australian continent, mostly from Western Australia. Results suggested an increase of 3 °C of mean ground surface temperature over mean annual air temperature. Since then, several Australian studies have adopted this +3 °C offset as a standard correction to apply to the mean annual air temperature as a method to estimate ground surface temperature (c.f. Beardsmore and Cull, 2001, Cull and Conley, 1983, Meixner et al., 2012). Similar corrections have also been applied internationally (e.g. Bartlett et al., 2006, Blackwell et al., 1980, Deming and Chapman, 1988, Majorowicz and Jessop, 1981).

Several previous studies have estimated the land-surface temperature at specific points across the Australian continent. In a local study of the Carnarvon Basin in Western Australia, Beardsmore (2005) extrapolated observed borehole thermal gradients to show that the estimated mean land-surface temperature was 6 °C warmer than mean annual air temperature; twice the average suggested by Howard and Sass (1964). Gerner and Budd (2015) analysed data from 108 continuous borehole temperature logs and 81 Bureau of Meteorology ground temperature sensors from across the continent and demonstrated that the difference compared to average annual air temperature was 3.38 °C averaged across the entire dataset. This result was not substantially different from that of Howard and Sass (1964), however, with individual sites recording differences of 1–8 °C, a fixed correction factor of 3 °C can result in errors of up to 5 °C in some areas. Unfortunately, Gerner and Budd (2015) lacked the spatial detail required to draw rigorous quantitative conclusions and estimates of uncertainty.

Remote sensing provides an opportunity to achieve the spatial resolution required for detailed mapping of land-surface temperature at regional and continental scales. Preliminary studies have previously demonstrated the potential for such an approach to be applied within a geothermal context (Horowitz and Regenauer-Lieb, 2009, Horowitz, 2015). In particular, Horowitz and Regenauer-Lieb (2009) presented a map of mean Australian land-surface temperature. Their study however, was limited by the six year record of satellite data available. The general approach developed in these earlier studies is adopted here with modification. The method is applied to the whole Australian continent, using a much longer record of observations, and with comparison to other independent measurements of mean land-surface temperature. The validity of an empirical +3 °C correction to the mean annual air temperature will also be briefly examined.

Section snippets

Remote sensing

Daily land-surface temperature observations are available as a standard data product of the Moderate Resolution Imaging Spectroradiometer (MODIS). MODIS sensors are installed on two satellite systems; Terra and Aqua. The two satellites are currently operational with Terra commencing observations on 5 March 2000, and Aqua commencing observations on 8 July 2002. The satellites are sun-synchronous, recording both a day-time and night-time pass for a given location in a single 24 h period. The orbit

Results and discussion

The mean annual land-surface temperature, as calculated from the average mean of the 13 annual periods calculated using (4), is shown in Fig. 3. The mean amplitude calculated across the 13 years using (5), representing the magnitude of the seasonal change in temperature, is shown in Fig. 4. The mean phase shift calculated across the 13 years using (6), being inversely proportional to the timing of the annual temperature peak, is shown in Fig. 5. Three-by-three covariance matrices were

Conclusions

Land-surface temperature is an important boundary condition used to constrain the thermal budget of the crust in geothermal studies. The use of annual average air temperatures, even with standard corrections, can result in underestimation of up to 9.2 °C in the approximation of localised land-surface temperatures. The new annual land surface temperature map presented in this paper, as derived from MODIS satellite data, provides a close match to other independent methods of estimating surface

Acknowledgements

Marcus Haynes thanks Geoscience Australia for a PhD scholarship to undertake this work. Ed Gerner publishes with the permission of the CEO, Geoscience Australia. The authors thank Richard Chopping, Stephen Sagar, and our two anonymous reviewers for their constructive feedback and advice.

References (48)

  • Australian Bureau of Meteorology

    Average daily mean temperature, Annual [Dataset]

    (2016)
  • M.G. Bartlett et al.

    Snow effect on North American ground temperatures, 1950–2002

    J. Geophys. Res. Earth Surf.

    (2005)
  • M.G. Bartlett et al.

    A decade of ground-air temperature tracking at emigrant pass observatory, Utah

    J. Climate

    (2006)
  • G.R. Beardsmore

    High-resolution heat-flow measurements in the southern Carnarvon Basin, Western Australia

    Explor. Geophys.

    (2005)
  • G.R. Beardsmore et al.

    Crustal Heat Flow: A Guide to Measurement and Modelling

    (2001)
  • G.R. Beardsmore et al.

    A protocol for estimating and mapping global EGS potential

    GRC Trans.

    (2010)
  • D.D. Blackwell et al.

    The terrain effect on terrestrial heat flow

    J. Geophys. Res.

    (1980)
  • D.S. Chapman et al.

    Heat flow in the Uinta Basin determined from bottom hole temperature (BHT) data

    Geophysics

    (1984)
  • C. Clauser

    Numerical Simulation of Reactive Flow in Hot Aquifers: SHEMAT and Processing SHEMAT

    (2003)
  • J.P. Cull

    An appraisal of Australian heat-flow data

    BMR J. Aust. Geol. Geophys.

    (1982)
  • J.P. Cull et al.

    Geothermal gradients and heat flow in Australian sedimentary basins

    BMR J. Aust. Geol. Geophys.

    (1983)
  • C. Danis et al.

    Gunnedah Basin 3D architecture and upper crustal temperatures

    Aust. J. Earth Sci.

    (2010)
  • G.F. Davies

    Dynamic Earth: Plates, Plumes and Mantle Convection

    (1999)
  • D. Deming et al.

    Heat flow in the Utah-Wyoming thrust belt from analysis of bottom-hole temperature data measured in oil and gas wells

    J. Geophys. Res.

    (1988)
  • Cited by (7)

    • Land use-land cover (LULC) transformation and its relation with land surface temperature changes: A case study of Barrackpore Subdivision, West Bengal, India

      2020, Remote Sensing Applications: Society and Environment
      Citation Excerpt :

      Thermal Infrared region (TIR) of satellite imageries can capture surface temperature, emissivity responses of an individual object over the surface which is very helpful to get the LST with high to low-resolution scale (Sobrino et al., 2008; Li et al., 2013; Lo et al., 1997). Landsat 5 TM; Landsat 7 ETM+; Landsat 8 (OLR) satellite data series have a thermal infrared region (10.40 μm −12.50 μm) from which the brightness temperature can calculate using the algorithms provided by the National Aeronautics and Space Administration(NASA) (Qin et al., 2001; Walawender et al., 2013; Wang et al., 2016) rather than other several geostationary earth observation satellite are also having TIR region to obtain Land surface temperature i.e., GOES (Geostationary Operational Environmental Satellite) with 4 km resolution, National Oceanic and Atmospheric Administration-Advanced Very High-Resolution Radiometer (NOAA-AVHRR) (Meng et al., 2019), Moderate Resolution Imaging Spectroradiometer (MODIS) of TERRA and AQUA with 1 km resolution (Haynes et al., 2018) and Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) of 90m resolution. In the worldwide (international field) several research works were done with LST increases in relation to land use changes.

    • Spatial structure of surface urban heat island and its relationship with vegetation and built-up areas in Melbourne, Australia

      2019, Science of the Total Environment
      Citation Excerpt :

      On this basis, considerable studies have attempted to understand the relationship and spatial distribution of LST, NDBI, and NDVI in various metropolitan areas worldwide (Zhou et al., 2016; Vargo et al., 2016; Sodoudi et al., 2014; Dos Santos et al., 2017; Chang, 2016; Fujibe, 2011). The results of this study show that the LST deals with the anomalies in different days of summer from 2014 to 2018, which is consistent with the findings of (McAlpine et al., 2007; Nicholls, 2006; Haynes et al., 2018). This study considers multiple days as acquisition dates in the summer season for LST analysis.

    View all citing articles on Scopus
    View full text