Return to AMS Hydrology Symposium Papers, Table of Contents
Return to Archive/Publications
SENSIBLE HEAT FLUX MEASUREMENTS USING A LARGE APERTURE SCINTILLOMETER OVER HETEROGENEOUS SURFACE
A. Chehbouni 1 *, O. Hartogensis 2, Y.H. Kerr 5 , L. Hipps 3, J.-P. Brunel 1, C.Watts 4,
J. Rodriguez 4, G. Boulet 1 , G. Dedieu 5 , and H. De Bruin 2
1 ORSTOM/IMADES, Hermosillo, Sonora, Mexico
2 WAU, Depart of Meteorology, Wageningen, The Netherlands
3 Utah State University, Logan, UT, USA
4 IMADES, Hermosillo, Sonora, Mexico
5 CESBIO, Toulouse, France
* Corresponding author: Reyes & Aguascalientes Esq., Col. San Benito, Hermosillo, C.P. 83190, Sonora, Mexico; email: email@example.com
The issue of estimating areal average of the turbulent surface fluxes over large and heterogeneous surfaces has been addressed in the framework of different field experiments. It is therefore well accepted that the above objective cannot be fulfilled without deploying a network of several single point measurements of surface fluxes, i.e., eddy correlation stations for example. Beside some of the technical limitations related to the required horizontal homogeneity of the surface layer, these systems are very expensive and required well trained staff to maintain (DeBruin et al. 1995).
Several investigations have recently demonstrate the potential of using scintillometer to measure areally averaged sensible heat fluxes over path lengths which are similar to satellite pixel scale i.e., several kilometers (Kohsiek, 1987; De Bruin et al., 1995; McAneney et al., 1995; Lagouarde et al., 1996; and Hartogensis, 1997). The idea behind the use of the scintillometer is based upon the consideration that the refractive index structure parameter measured directly by the scintillometer (C2n ) can be related to the structure function parameter of temperature (C2T) which is used to derive sensible heat flux. The objective of this study is to use the scintillometer to estimate areally averaged sensible heat flux over sparse grass in the Mexican part of the San Pedro basin. The scintillometer-based sensible heat flux over 3 different pathlengths (300, 600 and 900m ) is compared to that obtained using an eddy correlation method.
A scintillometer is an instrument that measure the intensity fluctuations of a light beam after propagating through a turbulent medium. It is assumed that these intensity fluctuations are caused by inhomogenities in the refractive index, which are due to turbulent eddy motions along the scintillometer path. These eddy motions are generated by temperature and humidity fluctuations, and can be regarded as a collection of converging and diverging lenses focusing and defocusing the scintillometer beam. (McAneney et al., 1996). In the visible and infra-red region, and assuming that temperature and humidity fluctuations are perfectly correlated, the spatially averaged refractive index structure parameter measured directly by a Large Aperture Scintillometer (LAS) is related to temperature structure parameter as:
where T is the air temperature (K) , P is the atmospheric pressure (Pa) and [beta] is the Bowen ratio (see DeBruin et al., 1995 for more details). The Bowen ratio term in Eq. (1) is an humidity correction term. The study by De Wekker (1996) has shown that this term can be neglected whenever the Bowen ration is greater than 0.6, which is generally the case in arid and semi-arid areas. Since the sensible heat and momentum fluxes together determine atmospheric stability, which in turn influences turbulent transport, an iterative procedure is needed to calculate sensible heat flux from the scintillometer measurement (Lagouarde et al., 1996). We first define the dimensionless temperature scale [theta]* as :
where [rho] is the density of the air and cp is the heat capacity at constant pressure, and u* is the friction velocity as :
where z is the measurement height , zo is the roughness length and d is the displacement height and [psi]m is the integrated stability function, and Lmon is the Monin-Obukhov length defined as:
Under unstable condition, Wyngaard et al. (1971) and DeBruin et al. (1995) found that the temperature structure parameter and [theta]* are related as :
Sensible heat flux can then be derived using Eqs (1-5) and an iterative procedure.
This study is a part the SALSA program (See Goodrich et al. , this issue). The objective of the investigation in the Mexican part of the Upper San Pedro basin is to better understand ecosystem function, and manage scarce natural resources by initiating the development and validation of a coupled SVAT and vegetation growth model for semi-arid regions that will assimilate remotely sensed data with several years of observed data.
Instrumentation were deployed during the summer of 1997 over sparse grass at the Zapata village situated about 10km East of Cananea (31,013° N, 110,09° W). The soil was mainly sandy loam. A Tower has been installed to measure conventional meteorological data (incoming radiation, net radiation, wind speed and direction , air temperature etc. etc.). Soil moisture was measured at both sites using Time Domain Reflectometry (TDR) sensors and Theta probes with sensitivity to 5 cm depth. Surface temperature was measured at two different view angles (0 and 45 degrees). Multidirectional ground based surface reflectance and the RED and NIR wavebands has been taken once a week during the entire growing season. Measurements of vegetation biomass, water content and leaf area index were made once a week. An eddy covariance system developed at the University of Edinburgh : Edisol (Moncrieff et al., 1997) has been used to measure turbulent surface fluxes. The system is made up of a three-axis sonic anemometer manufactured by Gill instrument ( Solent A1012R) and an IR gas analyzer (LI-COR 6262 model ) which is used in close path mode. The system is controlled by specially written software which calculates the surface fluxes of momentum , sensible and latent heat and carbon dioxide, from the output of the sonic and IR gas analyzer and displays them in real time. The software performs coordinate rotation on the raw wind speed data and allows for the delay introduced into CO2/H20 signal as a result of the time of the travel down the sampling tube.
The Large Aperture Scintllometer used in this study has been designed and built at the department of meteorology of the WAU. The electronics are according to Ochs and Wilson (1993). It has an aperture size of 0.15 m and the light source is a light emitting diode (LED: TIES-16A, Texas Instruments, Texas, USA) operating at a peak wavelength of 0.94 mm, which is placed at the focal point of a concave mirror. The reflected beam emitted by the transmitter diverges slightly at about ~0.002°. The irradiance distribution over the beam is completely uniform. The receiver also employs an identical mirror to focus light on a photo diode detector. To distinguish the light emitted by the LAS from ambient radiation it is excited by a 7 kHz square wave. Scintillations appear as amplitude modulations on the carrier wave. Receiver and emitter are both installed on tripods at a height of 2.6 m. For beam alignment, telescope rifle sights are mounted on both the receiver and emitter housings. The receiver electronics have been designed in such a way that after setting the path length , it gives an output voltage from which C2n follows using with Vout is the output voltage of the scintillometer. 10 minute values of the scintillometer output (Vout) were stored on a data logger (Campbell Scientific Ltd, 21X). After downloading of these data, they were linearly averaged to 30-minute averages of Vout.
4. DATA ANALYSIS
The experiment took place in three phases. First, the LAS was installed over a transact of about 300m centered around the tower where the Solent was installed. The choice of 300m pathlength was meant to make sure that the LAS and the EC are integrating the same surface. Comparison between sensible heat flux values measured with the Solent (HSolent) and that obtained by the scintillometer (HLAS) from Day Of the Year (DOY) 206 to DOY 216 under unstable conditions is presented in Figure 1. This figure shows that the two measurements are in very good agreement especially with a range of sensible heat values up to 450 Wm-2. A linear regression forced to the origin yields HLAS=0.97 HSolent , with a correlation coefficient R2= 0.90.
During the second phase, the receiver and the transmitter were moved 150m in the opposite direction, which make the pathlength of about 600m. Figure 2, presents the same comparison as in Figure 1 for data taken under unstable condition from DOY 240 to DOY 250. It is seen that although the agreement is still good, the scatter is larger than in Figure 1, the correlation coefficient drops to 0.80.
During the third phase, the receiver and the transmitter were moved so that the pathlength was about 900 m. In Figure 3, the path-averaged values of sensible heat flux obtained by the LAS from DOY 258 to DOY 261 is plotted versus the corresponding values measured with the Solent.
It can be seen here that scatter between the two measurements is getting larger, the correlation coefficient drops to 0.75. It is therefore clear that the scatter between the LAS and the Solent measurements increases when difference between the Solent footprint and the LAS pathlength increases. To emphasize this point, we present in Figure 4 the spatial variation of a vegetation index (MSAVI, Qi et al., 1994) derived from surface reflectance measurements using a field radiometer (Exotech) along the pathlength. The MSAVI value varies from 0.15 to 0.30 along the transact.
These variations of MSAVI values along the transact which can be associated to variation of surface albedo and temperature and therefore to sensible heat flux. This rises the issue of the reliability of point based measurements of surface fluxes to validate large scale models even when the surface looks homogeneous.
5. CONCLUDING REMARKS
Several successful studies have been carried to investigate the feasibility of estimating sensible heat flux using the scintillometer (Kohsiek 1985; Green et al., 1994; De Bruin et al. 1995; McAneney et al. 1995, Lagouarde et al., 1996). This study confirms the results reported elsewere and stressed out the problem of the non-representativeness of point measurements even over a quasi-homogeneous surfaces.
The LAS has several major advantages over more traditional methods, such as eddy correlation, for measuring sensible heat flux: a) it gives statistically more reliable data; b) it is not sensitive to flow distortions near the instrument; c) it is relatively cheap and easy to operate. Additionally, point based estimation of surface fluxes which are based on Monin-Obukhov similarity theory that requires horizontal homogeneity which is not always met in real world situation. The scintillometer can provides areally averaged values of sensible heat flux over heterogeneous surfaces by using the concept of blending height . This concept assumes that on a certain height above the surface, the differences in surface properties can be averaged out so that the similarity theory can be applied.
We acknowledge financial support from CONACYT, the French PNTS and the European Space Agency through VEGETATION and ERS2/ATSR2 projects. This research is situated within the framework of NASA Mission to Planet Earth (MTPE): NASA/EOS grant NAGW2425. Additional support was also provided by the USDA-ARS Global Change Research Program, NASA grant W-18,997, NASA Landsat Science Team, grant #S-41396-F.
De Bruin, H.A.R., Kohsiek, W. and van den Hurk, B.J.J.M. 1993: A verification of some methods to determine the fluxes of momentum, sensible heat, and water vapour using standard deviation and structure parameter of scalar meteorological quantities. Boundary layer meteorology 63, 231-257.
De Bruin, H.A.R., van den Hurk, B.J.J.M. H.A.R. and Kohsiek, W. 1995: The scintillation method tested over a dry vineyard area. Boundary layer meteorology 76, 25-40.
De Wekker, S.F.J. 1996: The estimation of areally-avaraged sensible heatfluxes over complex terrain with a Large-Aperture Scintillometer. M.S. thesis. Dept. of Meteorology, Wageningen Agricultural University. 42pp.
Goodrich D.C. et al., An overview of the 1997 activities of the Semi-Arid Land-Surface Program (this issue).
Green, A.E., McAneney, K.J. and Astill, M.S. 1994: Surface-layer scintillation measurements of daytime sensible and momentum fluxes. Boundary layer meteorology 68, 357-373.
Hartogensis Oscar. 1997: Measuring areally-averaged sensible heat fluxes with a Large Aperture Scintillometer: M.S. Thesis, Department of Meteorology Agricultural University of Wageningen the Netherlands
Kohsiek, W. 1985: A comparison between line-averaged observations of from scintillation of a CO2 laser beam and time averaged in situ observations. J. Clim. and Appl. Meterol. 24, 1099-1102.
Kohsiek, W. 1987: A 15 cm aperture LED scintillometer for and crosswind measurements. KNMI Scientific Reports WR 87-3.
Lagouarde, J-P, McAneney, K.J. and Green, A.E. 1996: Scintillometer measurements of sensible heat flux over heterogeneous surfaces. in: Scaling up in hydrology using remote sensing J.B. Stewart, E.T. Engman, R.A. Feddes and Y. Kerr (eds.). Institute of Hydrology.
McAneney, K.J., Green, A.E. and Astill, M.S. 1995: Large aperture scintillometry: The homogenous case. Agricultural and Forest Meteorology 76, 149-162.
Moncrieff J.B., Massheder J.M., De Bruin H., Elbers J., Friborg T., Heusinkveld , Kabat P., Scottt, Soegaard H., and Verhoef A. (1997). A system to measure surface fluxes of momentum, sensible heat, water vapor and carbon dioxide, J of Hydrology, vol 188-189, pp 589-611.
Ochs, G.R. and Wilson, J.J., 1993: A second-generation large-aperture scintillometer. NOAA Tech. Memo, ERL WPL-232. NOAA Environmental Research Laboratories, Boulder, Co.
Qi, J., Chehbouni, A., Huete, A.R., Kerr, Y.H., and S. Sorooshian, 1994, A Modified Soil Adjusted Vegetation Index, Remote Sens. Environ., Vol 48, N°2, pp 119-126.
Return to AMS Hydrology Symposium Papers, Table of Contents
Return to Archive/Publications