FLUX AGGREGATION OVER HETEROGENEOUS SURFACES
IN A SEMI-ARID REGION : MODELLING AND VALIDATION
IN A SEMI-ARID REGION : MODELLING AND VALIDATION
Y.H. Kerr (1)*, A. Chehbouni (2), L.H. Hipps (3), J. Schieldge (4), D.I. Cooper (5), W.E. Eichinger (6), R. Scott (7), D. C. Goodrich (8), J. Shuttleworth (7), K.H. Lee (7), C. Watts (9).
(1) CESBIO (CNES/CNRS/UPS), Toulouse, France; (2) ORSTOM/IMADES, Hermosillo, Sonora, Mexico; (3) Utah State Univ., Logan, UT; (4) Jet Propulsion Lab., Pasadena, CA, USA; (5) Los Alamos Nat. Lab., Los Alamos, NM, USA; (6) Univ. of Iowa, Iowa City, IA USA; (7) University of Arizona, Tucson, AZ, USA; (8) USDA-ARS, Tucson, AZ, USA; (9) IMADES, Hermosillo, Sonora, Mexico. * Corresponding author: CESBIO BPI, 18 Ave. E. Belin, 31401, Toulouse, Cedex 4, FRANCE; email: Yann.Kerr@cesbio.cnes.fr
1. INTRODUCTION
The primary objective of the SALSA Program is to understand, model and predict the consequences of natural and human-induced change on the basin-wide water balance and ecological diversity of semiarid regions at event, seasonal, interannual, and decadal time scales (see Goodrich et al., 1998 ; this issue). The scientific objectives of the Salsa program are described in Goodrich et al., 1994. Among the scientific objectives there are the will to:
1. Improve the diagnosis of surface fluxes used in atmospheric models with grid spacings of several kilometers and compare remote and in-situ observations with real-time model runs, and ; 2. Develop and validate aggregation schemes with data taken over very highly heterogeneous surfaces.
To achieve these goals it is necessary to be able to model how to translate elementary fluxes (homogeneous patches) into aggregated fluxes compatible with model scales. This task requires to cope with two problems.
The first one is to model the aggregation issue as the problem is not linear. The second is to validate the results. In this presentation we will show how we coped with these two issues and the very first results gained from the SALSA ’97 campaign.
2. DATA SETS
2.1 THE STUDY AREA
The Upper San Pedro Basin (USPB) was identified as the focal area for initial SALSA research during the 1995 workshop noted above. In addition to those factors mentioned in the prior section, the basin embodies a number of characteristics which make it an exceptional outdoor laboratory to address a large number of scientific challenges in arid and semi-arid hydrology, meteorology, ecology, and social and policy science. The basin represents a transition area between the Sonoran and Chihuahuan deserts. It is an international basin spanning the Mexico-United States Border with significantly different cross border legal and land use practices, significant topographic and vegetation variation, and a highly variable climate.
The annual rainfall ranges from around 300 mm to 750 mm with the majority
of annual precipitation (~65%) occurring during the July through September
monsoon season from high intensity air-mass convective thunderstorms and
roughly 30% coming from less intense winter frontal systems. Potential
evapotranspiration is high, estimated at over ten times annual rainfall.
Major vegetation types include desert shrub-steppe, grasslands, oak savannah,
pinyon-juniper, and ponderosa pine. In portions of the basin all of these
vegetation types are contained within a 20 km span.
As we had to start working on the aggregation issue before all of the measurements
were performed and available, we started to work with the HAPEX SAHEL data
set. Hapex Sahel took place in 1992 in Niger (see HAPEX SAHEL Journal of
Hydrology special issue, 1997). The area has a very similar climate to
that of the USPB. The vegetation is very patchy but with a limited variety
of land use. The area is essentially flat and the rainfall pattern is concentrated
during the « monsoon » period which extends from mid June to
end of September.
2.2 MEASUREMENTS
HAPEX SAHEL :
The measurements performed during Hapex Sahel are described in Prince et al., 1995. They consisted mainly in characterisation of the surface components (soils, vegetation type and density), measurements energy and water fluxes (rainfall, storage [ground water and ponds], sensible and latent heat fluxes, net radiation and ground flux, air and surface temperature, soil moisture profiles, etc...). The ground measurements were performed mainly over three so called « supersites ». Meteorological measurements and radiosoundings were done at several locations over the square degree area. Aircraft measurements were also performed during the intensive observation period (remote sensing and fluxes). Finally satellite acquisitions were collected on an almost routine fashion. An overview of the data collected can be found on the WEB site « http://www.orstom.fr/hapex/index.htm ».
SALSA ’97 :
Experimental surface observations were obtained from a variety of existing
as well as newly initiated data collection activities that were integrated
into the 1997 SALSA Program activities. We worked mainly with the Lewis
Spring site in the USA data as well as with the Zapata site in Mexico (see
Goodrich et al., 1998 this issue). The sites have sparse desert grass cover
and a wide variety of in-situ and remotely sensed data. These measurements
include basic meteorology, energy and CO2 fluxes, soil moisture, vegetation
sampling, and surface reflectances/emittance. We worked mainly with the
data collected during the August 1997 campaign, when additional instrumentation
deployed included an array of eddy correlation flux instrumentation (see
ET summary by Hipps et al., this issue), the scintillometer, and the Las
Alamos National Laboratory Raman LIDAR system (Cooper et al., this issue).
The main innovation in the measurements consisted in an attempt to use
scintillometry to quantify fluxes over scales compatible with models (McAnnaney
et al., 1995 ; Green et al., 1997) .
3. METHODS
3.1 MODELLING
3.2 SCINTILLOMETER
4. PRELIMINARY RESULTS
Work presented in this paper corresponds only to very preliminary results.
Actually, we simply tested the approach on the HAPEX SAHEL data set and
intend to validate it more extensively on the SALSA data where we will
have access to scintillometer measurements. It must also be stated that
a full validation will require the analysis of data collected over an area
with contrasted conditions.
To test the results we compared the surface temperature computed with the
above mentioned approach to the surface averaged temperature. The results
are given below in table 1. The fourth column gives the average temperature
difference between the two methods, the fifth the standard deviation and
the last one the maximum value encountered, all are expressed in percentage
with regard to the surface averaged temperature.
One can see that the temperature difference is not significantly high (2 to 3 %) which correspond to the rather uniform surface conditions. However, it can reach high values (13% which amounts to 5 K).
The same comparison method was applied to the sensible heat flux, and is depicted on figure 2. Again the average is not very high (5%) but we found a high standard deviation (7%), and correspondingly, high values for the extreme differences, which can reach 30% of the flux.
Finally, we compared the composited and surface averaged latent heat fluxes (table 3). If the mean value of the difference is very close to that of the sensible heat flux, the standard deviation is much lower as for the maximum difference.
This study shows that the simple surface averaged temperature and fluxes may be significantly different. The next question is thus to identify which one is closest to reality. For this we investigated the use of the AVHRR for temperature and MERLIN data for fluxes. With the AVHRR we did not have much success since the dates when we had all the ground measurements were often cloudy at the time of overpass. Secondly we had only access to averaged surface temperatures and it was difficult to relate them to the instantaneous AVHRR measurement. The aircraft acquisitions gave us also mixed feelings. It seems that there is a constant overestimation (whichever the method, including measured fluxes) of ground fluxes vs aircraft fluxes. Consequently the uncertainties do not allow to draw any conclusions as the A/C fluxes are way out of the results from either the model or the surface averaged fluxes. We intend to use the sensible heat flux measured with the scintillometer to verify our model.
Finally we checked the sensitivity of the two approaches to the accuracy of the surface cover estimates. For this we changed the percentage of cover for millet and fallow and compared the results as shown in table 4. Taking as a reference an area with 25% millet, 25 % fallow and 50 % bare soil, we changed the proportions to 8, 17,25, 33,42 % for each vegetation component. The bare soil proportion varying accordingly. The number corresponds to the change of the considered parameter in percent. The results shown are the means for 9 days (same days as in table 1 to 3). As expected in such environment, the most important effect is that of bare soil for temperature. However the mean variation is less than 2 °C on average, but it can reach 4°C for maximum difference and extreme error in surface cover. For sensible and latent heat fluxes, the role of surface cover is somewhat larger as it can reach 27% change. Moreover it can be seen on the tables that for the same amount of bare soil, the respective contributions of fallow and millet are no more negligible.
In conclusion, one may say, that even if an extreme accuracy is not required, a good knowledge of surface cover is nevertheless required.
5. CONCLUSION
This paper presents the very first results from our attempt to model aggregation. We found that there were differences between our model and an area weighted average. We also found that with our model, as expected, surface cover had to be known. Unfortunately, it is not yet possible from the data we had to decide which was the most accurate method. We believe that with the data collected last august on the ground and with the scintillometer, will enable us in the near future to select the most appropriate method and validate it.
6. ACKNOWLEDGMENTS
Financial support from the USDA-ARS Global Change Research Program, USDA National Research Initiative Grant Program, Electrical Power Research Institute, Arizona Department of Water Resources, U.S. Environmental Protection Agency -Office of Research and Development, CONACYT, ORSTOM, the French Remote Sensing Program (PNTS) via the VEGETATION Project and the ERS2/ATSR2 Project, and US Bureau of Land Management is gratefully acknowledged. Assistance was also provided in part by the NASA/EOS grant NAGW2425, EPA STAR Graduate Student Fellowship Program, National Science Foundation, US Geological Survey, US Department of Energy contract W-7405-ENG-36, California Institute of Technology-Jet Propulsion Laboratory (NASA, EOS/ASTER), WAU (Wag. Agricultural University, Netherlands); this support is also gratefully acknowledged. Special thanks are extended to the ARS staff located in Tombstone, Arizona for their diligent efforts. This work was performed in the framework of the French team effort for EOS/MTPE. We are very thankful to J.P. Lagouarde and J. McAneney for sharing with us their large expertise on scintillometry.
7. REFERENCES
Chehbouni A. et al., 1995. Approaches for averaging surface parameters and fluxes over heterogeneous terrain. Journal of climate. Vol 0. P 1-8.
De Bruin H.A.R., Van Den Hurk B.J.J.M., Kohsiek W., 1995 : The scintillation method tested over a dry wineyard area. Boundary-Layer Meteorol., 76, 25-40.
Goodrich, D. C., 1994: SALSA-MEX: A Large Scale Semi-Arid-Land-Surface-Atmospheric Mountain Experiment. Proc. 1994 Intern. Geoscience and Remote Sensing Sym. (IGARSS'94), Pasadena, CA, Vol. 1, Aug. 8-12, p. 190-193.
Green A.E., McAneney K.J., Lagouarde J.P., 1997 : Sensible heat and momentum flux measurement with an optical inner scale meter. Agric. For. Meteorol., 85, 259-267.
Hill R.J., Ochs G.R., Wilson J.J., 1992 : Surface-layer fluxes measured using the CT2-profile method. J. of Atmospheric and Oceanic Technology, 9, 5, 526-537.
Journal of hydrology, Special issue on Hapex Sahel, guest Eds Goutorbe, Dolman, Gash, Kerr, Lebel, Prince, Stricker, 188-189, no. 1-4.
Kohsiek W., 1982 : Measuring CT2, CQ2 and CTQ in the unstable surface layer, and relations to the vertical fluxes of heat and moisture. Boundary-Layer Meteorol., 24, 89-107.
McAneney K.J.,et al., 1995 : Large-aperture scintillometry
: the homogeneous case. Agric. For. Meteorol., 76, 149-162.
Prince, S.D.et al., 1995, Geographical, Biological and Remote Sensing Aspects
of the HAPEX SAHEL: Remote Sens. Envir., v. 51, p. 215-234.
Wyngaard J.C., 1973 : On surface-layer turbulence.Workshop on Micrometeorology, Denver, Colorado, Amer . Meteor. Soc., 101-149.
REFERENCES IN THIS ISSUE
Chehbouni, A., O. Hartogensis, L. Hipps, J-P. Brunel, C. Watts, J. Rodriguez, G. Boulet, G. Dedieu, Y. Kerr, and H. De Bruin, (a) Sensible heat flux measurements using a large aperture scintillometer over heterogeneous surfaces.
Cooper, D.I., W. Eichinger and L. Hipps, Spatial properties of water vapor scalar and flux over a riparian corridor.
Goodrich, D.C. et al, An overview of the 1997 activities of the Semi Arid Land Surface Atmosphere (SALSA) program.
Hipps, L.E., D. Cooper, W. Eichinger, D. Williams, S. Schaeffer, K. Snyder, R. Scott, A. Chehbouni, C. Watts, O. Hartogensis, J-P.
Lhomme, B. Monteny, J-P. Brunel, G. Boulet, J. Schieldge,
H. De Bruin, J. Shuttleworth, and Y. Kerr, A summary of processes which
are connected to evaporation of riparian and heterogeneous upland vegetation
in arid regions.
homepage