Research paper
Infrared surveillance of crater lakes using satellite data

https://doi.org/10.1016/0377-0273(93)90093-7Get rights and content

Abstract

Under optimum circumstances, thermal infrared data recorded from satellites can measure water surface temperatures to accuracies of a few tenths of a degree Celsius. Such techniques are applied here to evaluate volcanic crater lake temperatures. At present, band 6 of the Landsat Thematic Mapper (TM) is the most pertinent sensor in this respect, although its nominal 120 × 120 m “footprint” only permits useful measurements of circular lakes exceeding 340 m in diameter. In addition, the radiative properties of the atmosphere between sensor and target at the instant of observation should be well-characterised in order to make confident measurements of surface temperatures with single-band infrared data.

An analysis is presented of three TM band-6 images of the crater lake at Poás volcano, Costa Rica, recorded on February 6, 1986, March 13, 1987, and May 10, 1988. In the February 1986 scene, the band-6-derived water surface temperature is 36°C which is 3°C less than a field measurement made in the same month. Since the satellite measurement was integrated over some 14,400 m2 of the lake surface, while the field measurement was obtained at a single point below the surface, the former may be more representative of surface heat losses. Subsequent TM data reveal an increasing discrepancy between contemporaneous field and satellite observations, probably because the lake diameter had decreased to 250 m by March 1987, and to less than 200 m by mid-1988, greatly reducing the likelihood of obtaining a pure “lake” pixel.

The Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) which is earmarked for orbit on the first of NASA's Earth Observing System satellite platforms later this decade has five discrete spectral bands in the thermal infrared region and will produce data composed of 90 × 90 m pixels. These specifications could enable the determination of water surface temperatures of > 250 m diameter crater lakes by algorithms that implicitly account for atmospheric effects.

References (40)

  • L.A. Bartolucci et al.

    Atmospheric effects on Landsat TM thermal IR data

    IEEE Trans. Geosci. Remote Sens.

    (1988)
  • S.L. Brantley et al.

    Poás volcano crater lake acts as a condenser for acid metal-rich brine

    Nature

    (1987)
  • G. Brown et al.

    Energy budget analysis for Poás crater lake: implications for predicting volcanic activity

    Nature

    (1989)
  • G. Brown et al.

    Volcano monitoring by microgravity and energy budget analysis

    J. Geol. Soc. London

    (1991)
  • T.J. Casadevall et al.

    Sulfur dioxide and particles in quiescent plumes from Poás, Arenal and Colima volcanoes, Costa Rica and Mexico

    J. Geophys. Res.

    (1984)
  • D.E. Gibbons et al.

    Application of Landsat Thematic Mapper data for coastal thermal plume analysis at Diablo Canyon

    Photogramm. Eng. Remote Sens.

    (1989)
  • A.R. Harris et al.

    Lake area measurement using AVHRR

  • A.W. Hurst

    Temperature observations of Crater Lake, Mt Ruapehu, New Zealand, using temperature telemetry buoys

    Bull. Volcanol.

    (1980)
  • A.B. Kahle et al.

    The Advanced Spaceborne Thermal Emission and Reflectance Radiometer (ASTER)

    Int. J. Imaging Systems Tech.

    (1991)
  • F.X. Kneizys et al.

    Users' guide to LOWTRAN7

  • Cited by (43)

    • Thermal evolution of the Crater Lake of Copahue Volcano with ASTER during the last quiescence period between 2000 and 2012 eruptions

      2020, Journal of Volcanology and Geothermal Research
      Citation Excerpt :

      Thus, thermal and chemical monitoring of ACL makes possible the identification of precursory signals of volcanic activity that can provide information useful for volcanic eruptions forecasting (Rowe et al., 1992a, 1992b; Oppenheimer, 1997a; Varekamp et al., 2001; Dehn et al., 2002; Hernández et al., 2007; Colvin, 2008; Spampinato et al., 2010). Remote sensing has been widely used since the 1980s in order to investigate volcanic systems (Harris, 2013; Pyle et al., 2013) as well as features of ACL such as morphology, watercolor, temperature and heat flux (Oppenheimer, 1993, 1996, 1997a, 1997b; Colvin, 2008; Trunk and Bernard, 2008; Harris, 2013; Jay et al., 2013; Murphy et al., 2018). The observation of ACL with infrared instruments enables us to study low temperature phenomena or anomalies (Flynn et al., 2000), since although the lakes constitute the end of the conduit of an active volcanic system and they maintain the temperatures above ambient values, these do not generally exceed 100 °C.

    • Spatial heterogeneity in geothermally-influenced lakes derived from atmospherically corrected Landsat thermal imagery and three-dimensional hydrodynamic modelling

      2016, International Journal of Applied Earth Observation and Geoinformation
      Citation Excerpt :

      Thus for robust earth surface temperature estimations, atmospheric profile data are required. Satellite thermal imagery has been used for a wide variety of lake applications including: geologic exploration (van der Meer et al., 2014), temperature monitoring of volcanic (Oppenheimer, 1993) and other lakes (Alcântara et al., 2010; Giardino et al., 2001; Lamaro et al., 2013; Sima et al., 2013; Simon et al., 2014); characterising upwelling and circulation (Steissberg et al., 2005a); observing surface current speed/direction (Steissberg et al., 2005b) and near-shore thermal bars (Schott et al., 2001); identification of groundwater discharge areas (Tcherepanov et al., 2005); estimating the influence of lake morphology and clarity on water surface temperature (Becker and Daw, 2005); and validation of physically based 3-D hydrodynamic models (Pahlevan et al., 2012). Remote sensing radiometers measure the lake water ‘skin’ temperature (upper 100 μm) rather than the bulk surface temperature usually measured with in situ sensors.

    • Temporal variations in fumarole gas chemistry at Poás volcano, Costa Rica

      2015, Journal of Volcanology and Geothermal Research
      Citation Excerpt :

      Poás is an example of a volcano with a dynamic lake-hydrothermal system and lake parameters such as size, level, temperature, and composition are affected by volcanic activity (Brown et al., 1989; Martínez et al., 2000; Mora-Amador et al., 2010; Ramirez et al., 2010; Alpízar et al., 2014). Over the past three decades, monitoring of lake temperature (Oppenheimer, 1993; Vaselli et al., 2003; Trunk and Bernard, 2008), fluid composition (Rowe et al., 1992; Rowe, 1994; Martínez et al., 2000; Vaselli et al., 2003; Zimmer et al., 2004; Martinez, 2008; Hilton et al., 2010), plume chemistry (Pfeffer et al., 2006), and geophysical parameters (Casertano et al., 1987; Brown et al., 1989; Rymer and Brown, 1989) have provided critical insights into constraining processes occurring at Poás. Here, we show that key major gas ratios changed during 2005 to 2006, preceding a period of increased phreatic activity that began in March 2006 and continues to date.

    • Past, present and future of volcanic lake monitoring

      2014, Journal of Volcanology and Geothermal Research
    • Thermal deconvolution: Accurate retrieval of multispectral infrared emissivity from thermally-mixed volcanic surfaces

      2014, Remote Sensing of Environment
      Citation Excerpt :

      However, because of the limited spatial and spectral resolution of those data (1 km), the authors did not account for multiple high temperature components nor the surface emissivity. Previous studies have applied the dual band approach to volcanic processes using AVHRR and Landsat Thematic Mapper (TM) sensors to produce a relatively accurate estimation of temperature, radiant flux, and the behavior of volcanic features such as lava flows and fumaroles (Abrams, Glaze, & Sheridan, 1991; Flynn, Mouginis-Mark, & Horton, 1994; Glaze, Francis, & Rothery, 1989; Harris et al., 1998; Oppenheimer, 1993; Oppenheimer, Francis, Rothery, & Carlton, 1993; Oppenheimer & Rothery, 1991; Pieri et al., 1990; Rothery et al., 1988; Vaughan et al., 2010). However, inherent errors such as daytime solar reflected radiance, the presence of steam and gas plumes, precipitate minerals accumulated on the surface, topography relative to the viewing angle, and the width of each spectral channel of the instrument all made accurate thermal deconvolution difficult.

    View all citing articles on Scopus
    View full text