Elsevier

Remote Sensing of Environment

Volume 106, Issue 3, 15 February 2007, Pages 285-304
Remote Sensing of Environment

Regional evaporation estimates from flux tower and MODIS satellite data

https://doi.org/10.1016/j.rse.2006.07.007Get rights and content

Abstract

Two models were evaluated for their ability to estimate land surface evaporation at 16-day intervals using MODIS remote sensing data and surface meteorology as inputs. The first was the aerodynamic resistance–surface energy balance model, and the second was the Penman–Monteith (P–M) equation, where the required surface conductance is estimated from remotely-sensed leaf area index. The models were tested using 3 years of evaporation and meteorological measurements from two contrasting Australian ecosystems, a cool temperate, evergreen Eucalyptus forest and a wet/dry, tropical savanna. The aerodynamic resistance–surface energy balance approach failed because small errors in the radiative surface temperature translate into large errors in sensible heat, and hence into estimates of evaporation. The P–M model adequately estimated the magnitude and seasonal variation in evaporation in both ecosystems (RMSE = 27 W m 2, R2 = 0.74), demonstrating the validity of the proposed surface conductance algorithm. This, and the ability to constrain evaporation estimates via the energy balance, demonstrates the superiority of the P–M equation over the surface temperature-based model. There was no degradation in the performance of the P–M model when gridded meteorological data at coarser spatial (0.05°) and temporal (daily) resolution were substituted for locally-measured inputs.

The P–M approach was used to generate a monthly evaporation climatology for Australia from 2001 to 2004 to demonstrate the potential of this approach for monitoring land surface evaporation and constructing monthly water budgets from 1-km to continental spatial scales.

Introduction

The ability to monitor evaporation1 from land surfaces is important for applications requiring spatially-resolved estimates of moisture availability over large areas continuously at weekly to monthly timescales. Examples of such applications include irrigation scheduling (e.g. Dodds et al., 2005), managing carbon, water and land resources (e.g. Meyer, 1999, Raupach, 2001), and risk assessments for bushfires, dust storms and flooding. Evaporation is a large component of the terrestrial water balance, so improving the accuracy of evaporation estimates will significantly reduce uncertainties in terrestrial water balance modelling and improve the quality of information used in these applications.

Hydrometeorologists have striven for decades to use the global coverage of satellite-based remote sensing to provide accurate estimates of evaporation at daily to weekly time scales and at fine spatial scales (100 to 103 m). These efforts have been hindered by two problems: firstly, that the quantities of interest, such as carbon and water fluxes and their associated stores, must be estimated indirectly using algorithms that relate measured radiances to, for example, leaf area index (Myneni et al., 2002), gross and net primary productivity (Running et al., 2004), vegetation indices (Huete et al., 2002) and land surface temperature (Wan et al., 2002). Secondly, one of the biggest impediments to global, multi-temporal satellite-based monitoring is the conflicting requirement for algorithms that are biophysically realistic yet simple enough for global parameterisation and implementation. Zhao et al. (2005) demonstrates this for the MODIS primary productivity products, while a global MODIS evaporation product remains elusive because no algorithm has yet been found that achieves the right balance between accuracy and simplicity.

We present a new approach to building a global land surface evaporation algorithm using optical/thermal satellite data. Our goal is to develop an observation model appropriate for global implementation and routine monitoring of landscape-scale evaporation at weekly to monthly timescales. In this paper we use data products from the MODIS Moderate Resolution Imaging Spectroradiometer on the polar-orbiting Terra satellite, which has a daily overpass at around 10:30 h local time. With these constraints, the following model attributes are required:

  • i)

    Model inputs and parameters must be routinely available at daily time and local space scales, for large regions such as continental Australia, and globally.

  • ii)

    The model needs to be robust, i.e. evaporation estimates are constrained by energy and mass conservation and have relatively low sensitivity to the input data and parameters.

  • iii)

    The model needs to be insensitive to constraints imposed by the once-daily overpass of the polar orbiting satellite and the necessary cloud screening and compositing procedures.

  • iv)

    The model needs to be validated using comparable evaporation measurements from a diverse range of bioclimates.

These objectives are consistent with the ultimate goal for Fluxnet (Baldocchi et al., 2001), which seeks to integrate flux and concentration measurements, remote sensing and land-surface modelling to yield a comprehensive global biosphere monitoring network (Running et al., 1999, Zhao et al., 2005). Fluxnet encompasses over 400 towers distributed across the globe, providing hourly measurements of carbon, water and sensible heat fluxes across a diverse range of ecosystems and climates for multiple years (Baldocchi et al., 2001; http://www.daac.ornl.gov/FLUXNET/fluxnet.html). The insights and constraints provided by the simultaneous measurement of these fluxes and their corresponding scalar fields ensures that Fluxnet provides an excellent data set for land surface model development and testing. Data from two Australian flux stations (Ozflux: http://www.dar.csiro.au/lai/ozflux/) are used in this paper to test two evaporation models: i) an aerodynamic resistance–surface energy balance model and ii) the Penman–Monteith (P–M) equation, where the required surface conductance is estimated from remotely-sensed vegetation indices (leaf area index and NDVI).

The plan of the paper is as follows: Section 2 presents the fundamental energy balance and evaporation equations that underpin the evaporation modelling approaches that are tested; Section 3 describes the micrometeorological flux measurements used to develop and test the models; and Section 4 evaluates the two modelling approaches. 5 Implementing P–M at regional scales, 6 Concluding comments then implement the successful model to determine monthly evaporation fluxes at 1-km resolution for the Australian continent. This demonstrates the potential to monitor monthly land surface evaporation at the regional-scale by combining surface-based meteorological measurements with MODIS remote sensing.

Section snippets

Modelling land surface evaporation

Energy partitioning at the surface of the earth is governed by the following three coupled equations:H=ρcpTsTaRa,λE=ρcpγeseaRa+Rs,A=RnGΔS=H+λE,where H, λE and A are the fluxes of sensible heat, latent heat and available energy, Rn is net radiation, G is soil heat flux; ΔS is the heat storage flux; Ts, Ta are the aerodynamic surface and air temperatures; es, ea are the water vapour pressure at the evaporating surface and in the air; Ra is the aerodynamic resistance, Rs is the surface

Flux measurement sites

Fluxes of sensible and latent heat used in the model evaluation were measured over two strongly contrasting ecosystems. The first is a wet/dry tropical savanna located in northern Queensland (Virginia Park, 19°53′00″S, 146°33′14″E, elevation of 200 m ASL), the other is a cool temperate, broadleaved forest in south east New South Wales (Tumbarumba, 35°39′20.6″S, 148°09′07.5″E, elevation of 1200 m ASL). Complete site descriptions, including the soil, rainfall climate and flux station

Climatology and surface properties

Time series of TsR and ND for the 7 × 7 km2 MODIS cutouts for Tumbarumba and Virginia Park are shown in Fig. 1. Gaps in the original record due to presence of persistent cloud cover during the wet season have been filled using linear interpolation, which may limit the validity of TsR for the Virginia Park site in the first wet season (Dec. 2001–Feb. 2002). The ND for Tumbarumba had a mean value of 0.80 for 2001–2003, with a fairly small annual amplitude ranging from a maximum of 0.89 (late

Using non-local meteorological forcing

The above results suggest that the P–M model is the most appropriate model but these results were achieved using locally measured meteorological forcing. The next step is to investigate model performance when these local measurements are replaced by inputs derived from calibrated functions, broader scale meteorological data and look-up tables for parameters that vary with vegetation type. Daily values of incoming solar radiation, temperature and humidity are available from a variety of sources,

Concluding comments

Our goal was to develop a global model for monitoring land surface evaporation at weekly to monthly timescales and at 1-km to continental spatial scales using surface meteorology and MODIS remote sensing. We compared the remote sensing approach that uses radiative surface temperature to estimate evaporation with the P–M model. The resistance–surface energy balance approach yielded implausible and unrealistic estimates of evaporation. This approach failed because of the difficulties in using 8-

Acknowledgements

The authors gratefully acknowledge the technical support given by Steve Zegelin, Dale Hughes and Mark Kitchen in the design, construction and maintenance of the flux stations, and to Christine Pierret for establishing the database used in processing the flux station results. We are also grateful to Faith Ann Heinsch for extracting the DAO meteorological fields for the Ozflux sites and the comments of Dr Damian Barrett (CSIRO Land and Water) on an earlier draft. This work was supported in part

References (60)

  • R. Leuning et al.

    Carbon and water cycles in two contrasting Australian ecosystems: Wet/Dry savannas and cool temperate Eucalyptus forest

    Agricultural and Forest Meteorology

    (2005)
  • T.R. McVicar et al.

    Estimating one-time-of-day meteorological data from standard daily data as inputs to thermal remote sensing based energy balance models

    Agriculture and Forest Meteorology

    (1999)
  • T.R. McVicar et al.

    Using covariates to spatially interpolate moisture availability in the Murray–Darling Basin: A novel use of remotely sensed data

    Remote Sensing of Environment

    (2002)
  • R.B. Myneni et al.

    Global products of vegetation leaf area and fraction absorbed PAR from year one of MODIS data

    Remote Sensing of Environment

    (2002)
  • J.M. Norman et al.

    Source approach for estimating soil and vegetation energy fluxes in observations of directional radiometric surface temperature

    Agricultural and Forest Meteorology

    (1995)
  • S.W. Running et al.

    A global terrestrial monitoring network integrating tower fluxes, flask sampling, ecosystem modeling and EOS satellite data

    Remote Sensing of Environment

    (1999)
  • C.B. Schaaf et al.

    First Operational BRDF, albedo and nadir reflectance products from MODIS

    Remote Sensing of Environment

    (2002)
  • T.J. Schmugge et al.

    Remote sensing in hydrology

    Advances in Water Resources

    (2002)
  • Z. Wan et al.

    Validation of land-surface temperature products retrieved from Terra Moderate Resolution Imaging Spectroradiometer data

    Remote Sensing of Environment

    (2002)
  • M. Zhao et al.

    Improvements of the MODIS terrestrial gross and net primary production global data set

    Remote Sensing of Environment

    (2005)
  • D.D. Baldocchi et al.

    FLUXNET: A new tool to study the temporal and spatial variability of ecosystem-scale carbon dioxide, water vapour and energy flux densities

    Bulletin of the American Meteorological Society

    (2001)
  • W. Brutsaert et al.

    Test of a sensible heat transfer parameterization for surfaces with anisothermal dense vegetation

    Journal of Atmospheric Science

    (1996)
  • F. Caparrini et al.

    Mapping of land–atmosphere heat fluxes and surface parameters with remote sensing data

    Journal of Hydrometeorology

    (2003)
  • B.J Choudhury et al.

    Estimating soil wetness using satellite data

    International Journal of Remote Sensing

    (1988)
  • A.C. Dilley et al.

    Operational AVHRR processing modules: Atmospheric correction, cloud masking and BRDF compensation

    CSIRO Atmospheric Research Paper

    (2000)
  • P. Dodds et al.

    A review of methods to estimate irrigated reference crop evapotranspiration across Australia

  • J.J. Finnigan et al.

    A re-evaluation of long-term flux measurement techniques: Part 1. Averaging and coordinate rotation

    Boundary - Layer Meteorology

    (2003)
  • R.R. Gillies et al.

    Thermal remote sensing of surface soil water content with partial vegetation cower for incorporation into mesoscale prediction models

    Journal of Applied Meteorology

    (1995)
  • F. Hall et al.

    Satellite remote sensing of surface energy balance: Success, failures, and unresolved issues in FIFE

    Journal of Geophysical Research

    (1992)
  • S.B. Idso et al.

    Thermal radiation from the atmosphere

    Journal of Geophysical Research

    (1969)
  • Cited by (0)

    View full text