i Use of AVHRR-Derived Spectral Reflectances to Estimate Surface Albedo across the Southern Great Plains Cloud and Radiation Testbed (CART) Site J. Qiu and W. Gao Environmental Research Division Argonne National Laboratory, 9700 S. Cass Ave., Argonne, Illinois 60439 Extended abstract for the preprint volume of the Ninth Conference on Atmospheric Radiation 2-7 January 1997 Long Beach, California J This work was supported by the U.S. Department of Defense, Strategic Environmental Research and Development Program (SERDP), through the Atmospheric Radiation Measurement (ARM)/Unmanned Aerospace Vehicle (UAV) Program; and by the U.S. Department of Energy, Office of Energy Research, Office of Health and Environmental Research, under contract W31-109-Eng-38, as part of research on dry air-surface exchange for the Atmospheric Chemistry Program' by a contractor of the U. S. Government under contract No. W-31-104ENG-38. Accordingfv, the U. S. Government retains a nonexclusive. royalty-free license to publish or reproduce the published form of this U. S. Government purposes.
DISCLAIMER This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, make any warranty, expressor implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disdosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, pnwss, or service by trade name, trademark, manufacturer, or otherwise does not necessan7y corrstitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.
Portions of this document may be illegible in electronic image products. Images are produced from the best available onghil document.
USE OF AVHRR-DERIVED SPECTRAL REFLECTANCES TO ESTIMATE SURFACE ALBEDO ACROSS THE SOUTHERN GREAT PLAINS CLOUD AND RADIATION TESTBED (CART) SITE J. Qiu' and W. Gao Environmental Research Division Argonne National Laboratory, Argonne, Illinois 1. INTRODUCTION Substantial variations in surface albedo across a large area cause difficulty in estimating regional net solar radiation and atmospheric absorption o f shortwave radiation when only ground point measurements of surface albedo are used t o represent the whde area. Information on spatial variations and site-wide averages of surface albedo, which vary with the underlying surface type and conditions and the solar zenith angle, is important for studies of clouds and atmospheric radiation over a large surface area. Satellite measurements can generate better average surface information, such as albedo, than ground point measurements can provide. Satellite measurements cover a large area with a high spatial resolution. Many efforts have been devoted t o the estimation of surface albedo from satellite observations. One popular approach is t o use statistical matching t o find an empirical relation between the narrowband reflectance and the broadband reflectance and t o use an empirical relation t o describe t h e dependence of bidirectional reflectance on solar zenith angle (Ranson et ai. 1991). This approach is simple, but it depends on preexisting knowledge about the atmospheric conditions, the underlying surface, and the solar zenith angle for determination of the coefficients in t h e empirical relation. In this study, we used a bidirectional reflectance model t o inversely retrieve surface properties such as leaf area index (LAI) and then calculated the bidirectional reflectance distribution by using the same radiation model. The albedo was calculated by converting the narrowband reflectance t o broadband reflectance and then integrating over * Corresponding author address: Jie Qui, Environmental Research Division, Argonne National Laboratory, Argonne, IL 60439; e-mail <jqiu@anl.goo. the upper hemisphere. A set of ground multiband spectral reflectance measurements was used t o examine the modeled albedo; the model was then applied t o satellite images for the Southern Great Plains (SPG) Cloud and Radiation Testbed (CART) site operated by the Atmospheric Radiation Measurement (ARM) program. 2. MODEL An automatic inversion technique (Gao and Lesht 1996, Qui e t al. 1996) coupled with a bidirectional reflectance model was used t o retrieve surface parameters. The model (Gao 1993) uses a relatively simple approach t o describe the first-order scattering and multiple scattering of solar beams within a plant canopy and the interaction of the canopy scattering with the reflectance from the soil surface for a horizontally uniform plant canopy. The model allows for f a s t computation o f bidirectional reflectance and provides results that agree well with measurements (Gao 1993). Because of the lack of information for initializing the leaf angle distribution index in the inversion model (Qu i et al. 1996) over such a large area, we used a slightly different approach in the inversion scheme; t h a t is, we used a combination of the red channel and the near infrared (NIR) channel t o retrieve surface parameters (rather than a single channel). This greatly reduced the dependence of the retrieved surface parameters on the initial leaf angle distribution. The spectral albedos were first integrated over the shortwave region (0.4-3.0 mm) with an appropriate narrowband-to-broadband conversion mechanism. The conversion was based on the shortwave radiation model developed by Slingo and Schrecker (1 982), where the broadband reflectance, R, is calculated by
where sza, vza, vaz are the solar zenith angle, view zenith angle, and view azimuth angle, respectively; R1 and R2 are the narrowband reflectances calculated from the bidirectional reflectance model using the inversely retrieved surface properties and the solar and view angles; and the constants w l and w2 are set t o 0.5 as suggested by Slingo and Schrecker (1 982). Saunders (1 990) found that the calculated surface albedo is not very sensitive t o the values of w l and w2 assumed. The surface albedo was then calculated by spherical integration of the broadband reflectance over the upper hemisphere. The advantage of this method is that we can incorporate the bidirectional reflectance effect in the calculation of the surface albedo by using an automatic adjustment t o local surface conditions, rather than an empirical relation that does not account for changes in surface conditions. The automatic inversion scheme works well with satellite data and is theoretically applicable t o any kind of surface condition. scheme. The estimated surface albedo was then compared with the albedo derived from measurements o f the upwelling and downwelling shortwave radiation from the corresponding SIROS stations. Figure 1 compares the modeled albedo and the observed albedo at the SIROS radiation station at extended facility EF1. The modeled albedo is in good agreement with the field observation, with the best agreement near the solar noon hour. The deviation of the modeled values from the observed values increases slightly with increasing solar zenith angle. The obvious deviation of the measured albedo from its main trend between 1 5 0 0 and 18:OO was caused by cloud cover during that period of time. 3. MEASUREMENTS AND RESULTS A hand-held multispectral radiometer (MSR87) was used as a mobile platform for measuring narrowband surface reflectances in eight wavebands ranging from visible t o NIR regions. To facilitate comparison of the ground measurements with the corresponding satellite measurements, we chose two wavebands near the wavebands used in the advanced very-highresolution radiometers (AVHRR) on board NOAA1 2 and NOAA-14 satellites, the red channel at 600-627 nm and the NIR channel at 797-829 nm. All MSR87 ground measurements were made a t nadir viewing angle with various solar zenith angles (depending on the time o f the measurement). Measurements were made with this mobile radiometer during June 24-28, 1996, at ten solar and infrared radiation observation stations (SIROS) across the CART site, where the SIROS systems operate on a long-term basis t o measure upwelling and downwelling short- and longwave radiation. All SIROS stations are located over grassland or pasture land where grass height vanes widely, from about 0.05 m t o 1.O m. The multispectral reflectance measurements taken with the MSR87 at each of these SIROS stations were used t o estimate the surface albedo with the model-inversion-based Figure 7. Comparison of modeled and measured surface albedos at extended facility EF7 of the CART sire. 03 02 0.15 0.1 0.1 0.15 02 025 Observed Albedo 03 figure 2. Relation between modeled and observed minimal albedo at ten extended facilities across the CART site.
I I I 0.15 0.20 0.25 Albedo Figure 3. Image of the modeled surface albedo for the entire CART site. Figure 2 compares the observed minimum albedos near midday a t each of the ten sites and the corresponding values of modeled minimum albedo. All of the modeled albedos are within 0.03 of the observed albedos. Because most of the SIROS stations are installed above various kinds of grassland, the albedo range is very limited. To obtain a surface albedo representative of the CART site, we further applied the model scheme t o the AVHRR data extracted for the entire site. Figure 3 shows the modeled albedo for the CART site, calculated a t solar noon (corresponding t o the minimum albedo). The
, ' *-, * AVHRR data were obtained from the Pathfinder data set, averaged for the period of July 1-10, 1996, with a spatial resolution of 8 km. The relatively high albe'do (0.21-0.28) in the central area of the CART site reflects the underlying bare soil surface a f t e r the wheat there was harvested. In contrast, the albedo of about 0.18 for the east and west parts of the CART site is reasonable for the grassland/woodland there because the plant canopy absorbs more solar radiation than bare soil. 4. CONCLUSIONS We have developed a scheme t o estimate surface albedo solely on the basis of remotely sensed bidirectional reflectances, with the surface parameters necessary for estimating surface albedo automatically inverted from the satellite observed bidirectional reflectances. By using retrieved surface parameters such as LAI, one can take into account the bidirectional effect of the reflectance in the estimation of the albedo for various surface covers, because different bidirectional reflectances resulting from different vegetation densities can be included automatically in the model calculation without using empirical assumptions. Ground spectral reflectance measurements made with a hand-held radiometer were incorporated into the scheme t o estimate the surface albedo. The modelestimated and observed albedos a t the SIROS stations are in general agreement. Applied t o the AVHRR data extracted for the CART site, the scheme gave an albedo in reasonable agreement with the underlying surface conditions, indicating its potential t o further analyze the spatial and temporal variations of albedo over the CART site or a larger area. 5. ACKNOWLEDGMENT This work was supported by the U.S. Department of Defense, Strategic Environmental Research and Development Program (SERDP), through the Atmospheric Radiation Measurement (ARM)/Unmanned Aerospace Vehicle (UAV) Program; and by the U.S. Department of Energy, Office of Energy Research, Office of Health and Environmental Research, under contract W-3 1 109-Eng-38, as part of research on dry airsurface exchange for the Atmospheric Chemistry Program. The authors thank Dr. R. L. Coulter and D. R. Allen for their useful advise and - discussions. 6. REFERENCES Gao, W., 1993: A simple bidirectionalreflectance model applied t o a tallgrass canopy. Remote Sens. EnLirm., 45, 209-224. Gao, W., and Lesht, B. M., 1996: Model inversion of satellite-measured reflectances t o obtain surface biophysical and bidirectional reflectance characteristics of grassland. Accepted by Remote Sens. Envirm Qiu, J., Gao, W., and Lesht, 6. M., 1996: Inverting optical reflectance t o estimate surface properties of vegetation canopies, submitted t o Int. J. Remote Sens. Ranson, K.J., Irons, J. R., and Daughtry, C. S. T., 1991: Surface albedo from bidirectional reflectance. Remote Sens. Environ., 3 5, 201 21 1. - Saunders, R., 1990: The determination of broad band surface albedo from AVHRR visible and near-infrared radiances, In&. J. Remote Sens., 11, 49-67. Slingo, A., and Schrecker, H. M., 1982: On the shortwave radiative properties of stratiform water clouds, Quart. J. Roy. Meteor, SOC., 108, 407-426.