Smoke, Clouds, And Radiation-Brazil Research Articles

Retrieval of Optical Depth for Heavy Smoke Aerosol Plumes: Uncertainties and Sensitivities to the Optical Properties

This paper is concerned with uncertainties in the Advanced Very High Resolution Radiometer (AVHRR)based retrieval of optical depth for heavy smoke aerosol plumes generated from forest fires that occurred in Canada due to a lack of knowledge on their optical properties (single-scattering albedo and asymmetry parameter). Typical values of the optical properties for smoke aerosols derived from such field experiments as Smoke, Clouds, and Radiation-Brazil (SCAR-B); Transport and Atmospheric Chemistry near the Equator-Atlantic (TRACE-A); Biomass Burning Airborne and Spaceborne Experiment in the Amazonas (BASE-A); and Boreal Ecosystem‐Atmosphere Study (BOREAS) were first assumed for retrieving smoke optical depths. It is found that the maximum top-of-atmosphere (TOA) reflectance values calculated by models with these aerosol parameters are less than observations whose values are considerably higher. A successful retrieval would require an aerosol model that either has a substantially smaller asymmetry parameter (g , 0.4 versus g . 0.5), or higher singlescattering albedo (v k 0.9 versus v , 0.9), or both (e.g., g 5 0.39 and v 5 0.91 versus g 5 0.57 and v 5 0.87) than the existing models. Several potential causes were examined including small smoke particle size, low black carbon content, humidity effect, calibration errors, inaccurate surface albedo, mixture of cloud and aerosol layers, etc. A more sound smoke aerosol model is proposed that has a lower content of black carbon (mass ratio 5 0.015) and smaller size (mean radius 5 0.02 mm for dry smoke particles), together with consideration of the effect of relative humidity. Ground-based observations of smoke suggest that for t , 2.5 there is an increasing trend in v and a decreasing trend in g with increases in t, which is consistent with the results of satellite retrievals. Using these relationships as constraints, more plausible values of t can be obtained for heavy smoke aerosol. The possibility of smoke‐cloud mixtures is also considered, which can also lead to high TOA reflectances. However, without measurements, the hypothesis can be neither accepted nor negated. The study demonstrates that without independent assessments of the optical properties, large uncertainties would be incurred in the retrieved values of optical depth for heavy smoke aerosol plumes.

GOES-8 and NOAA-14 AVHRR retrieval of smoke aerosol optical thickness during SCAR-B

Using the NOAA-14 1-km Advanced Very High Resolution Radiometer (AVHRR) and the Geostationary Operational Environmental Satellite (GOES-8) imager data, smoke aerosol optical thickness ( ) is retrieved over land during the Smoke, Clouds and Radiation-Brazil (SCAR-B) experiment in Brazil during August-September 1995. The satellite-retrieved values are then compared against ground-based sunphotometer derived values from the Aerosol Robotic Network (AERONET) program. Both the AVHRR and GOES-8 retrieved values are in excellent agreement with the AERONET derived values with linear correlation coefficients of 0.93. A single scattering albedo of 0.90 (at 0.67 w m) provides the best fit between the GOES-8 and AERONET values. The sensitivity of the retrieved to assumed surface albedo and aerosol single scattering albedo are also examined. A simple multi-spectral thresholding algorithm is used to separate smoke from other features from GOES-8 satellite imagery and regional maps of are provided. Our results show that the aerosol properties used in this paper are adequate to characterize biomass burning aerosols and can be used in studies that model the role of biomass burning on regional climate.

Analysis of smoke impact on clouds in Brazilian biomass burning regions: An extension of Twomey's approach

Satellite remote sensing of smoke aerosol‐cloud interaction during the recent Smoke, Clouds, and Radiation‐Brazil (SCAR‐B) experiment is analyzed to explore the factors that determine the magnitude of the cloud response to smoke aerosol. Analysis of 2 years worth of data revealed that the response is greatest in the north of Brazil where aerosol optical depth is smallest, and tends to decrease as one moves southward, and as aerosol optical depth increases. Saturation in this response occurs at an aerosol optical depth of 0.8 in 1987 and 0.4 in 1995. To explore the reasons for this, a framework is developed in which the satellite‐measured response can be compared to simple analytical models of this response and to numerical models of smoke aerosol‐cloud interaction. Three types of response are identified: (1) cloud droplet concentrations increase with increasing aerosol loading, followed by saturation in the response at high concentrations; (2) as in type 1, followed by increasing droplet concentrations with further increases in aerosol loading. This increase in droplet concentration is due to the suppression of supersaturation by abundant large particles, which prevents the activation of smaller particles. This enables renewed activation of larger particles when smoke loadings exceed some threshold; (3) as in type 1, followed by a decrease in droplet number concentrations with increasing aerosol loading as intense competition for vapor evaporates the smaller droplets. The latter implies an unexpected increase in drop size with increasing smoke loading. The conditions under which each of these responses are expected to occur are discussed. It is shown that although to first‐order smoke optical depth is a good proxy for aerosol indirect forcing, under some conditions the size distribution and hygroscopicity can be important factors. We find no evidence that indirect forcing depends on precipitable water vapor.

Sensitivity of off-nadir zenith angles to correlation between visible and near-infrared reflectance for use in remote sensing of aerosol over land

Cloud absorption radiometer (CAR) multispectral and multiangular data, collected during the Smoke, Clouds, and Radiation-Brazil (SCAR-B) Experiment, was used to examine the ratio technique, the official method for remote sensing of aerosols over land from the moderate resolution imaging spectroradiometer (MODIS) data, for view angles from nadir to 650 off-nadir. The strategy used is to first select a pristine, low aerosol optical thickness flight, and then to compute ratios of reflectance at 0.47 and 0.68 /spl mu/m to corresponding values at 2.20 /spl mu/m, separately for backward and forward scattering directions. Similarly, the authors analyzed data from high turbidity flights for comparison purposes. For both flights, they removed the effects of atmospheric absorption and scattering using 6S, a radiative transfer code, and then recomputed the ratios again for different values of aerosol optical thickness. Finally, they analyzed bidirectional reflection function (BRF) data to examine the dependence of the ratio technique on the relative azimuth angle. Results of this analysis show that a relationship between visible reflectance and near infrared (IR) reflectance exists for view angles from nadir to 400 off-nadir, and that simple parametric relationships can be derived.

Estimation of Surface and Top-of-Atmosphere Shortwave Irradiance in Biomass-Burning Regions during SCAR-B

Abstract Using in situ measurements of aerosol optical properties and ground-based measurements of aerosol optical thickness (τs) during the Smoke, Clouds and Radiation—Brazil (SCAR-B) experiment, a four-stream broadband radiative transfer model is used to estimate the downward shortwave irradiance (DSWI) and top-of-atmosphere (TOA) shortwave aerosol radiative forcing (SWARF) in cloud-free regions dominated by smoke from biomass burning in Brazil. The calculated DSWI values are compared with broadband pyranometer measurements made at the surface. The results show that, for two days when near-coincident measurements of single-scattering albedo ω0 and τs are available, the root-mean-square errors between the measured and calculated DSWI for daytime data are within 30 W m−2. For five days during SCAR-B, however, when assumptions about ω0 have to be made and also when τs was significantly higher, the differences can be as large as 100 W m−2. At TOA, the SWARF per unit optical thickness ranges from −20 to −60 W m−2 over four major ecosystems in South America. The results show that τs and ω0 are the two most important parameters that affect DSWI calculations. For SWARF values, surface albedos also play an important role. It is shown that ω0 must be known within 0.05 and τs at 0.55 μm must be known to within 0.1 to estimate DSWI to within 20 W m−2. The methodology described in this paper could serve as a potential strategy for determining DSWI values in the presence of aerosols. The wavelength dependence of τs and ω0 over the entire shortwave spectrum is needed to improve radiative transfer calculations. If global retrievals of DSWI and SWARF from satellite measurements are to be performed in the presence of biomass-burning aerosols on a routine basis, a concerted effort should be made to develop methodologies for estimating ω0 and τs from satellite and ground-based measurements.

Use of the Ångstrom exponent to estimate the variability of optical and physical properties of aging smoke particles in Brazil

In situ airborne measurements from the Smoke, Clouds and Radiation‐Brazil (SCAR‐B) study show that during aging over 1–4 days the physical and optical properties of smoke particles are correlated. Consequently, if one optical or physical property of the smoke particles is determined, other properties can be derived. This methodology is validated using multiwavelength Ångstrom exponents determined from the ground‐based Sun photometer measurements in SCAR‐B. It is shown that the Ångstrom exponent determined from Sun photometers for the wavelength intervals 339–437 nm and 437–669 nm are well correlated with particle size, single‐scattering albedo, and the backscatter ratio (r2>0.8). Therefore, when almucantar sky radiance data are not available and for remote sensing applications (such as MODIS), some of the uncertainties in the properties of smoke particles can be reduced by applying these relationships. Using this methodology, major oscillations were observed in smoke particle properties in Brazil on timescales of ∼5–15 days, resulting in variations of the volume median diameter and single‐scattering albedo of ±0.04 μm and ±0.05, respectively. In comparison, the mean value of the dry smoke particle volume median diameter and single‐scattering albedo over all of Brazil was 0.27 μm and 0.86, respectively. A daily cycle in smoke particle properties was also observed. The weekly and seasonal variability in the single‐scattering albedo is shown to have significant consequences for retrieving aerosol optical depths from satellite measurements.

Correlation between smoke and tropospheric ozone concentration in Cuiabá during Smoke, Clouds, and Radiation‐Brazil (SCAR‐B)

Ozone soundings launched from Cuiabá between August 16 and September 10, 1995, during the Smoke, Clouds, and Radiation‐Brazil (SCAR‐B) experiment show an enrichment of tropospheric ozone when compared with average wet season values and also present a great variability in concentrations depending on the dominant circulation pattern. Smoke tracers, such as aerosol optical thickness, measured from a Sun photometer installed at the Instituto Nacional de Pesquisas Espaciais‐Cuiabá site, and black carbon ground measurements, also show an enhancement of smoke during the same period. Although there is a connection between the enrichment of the tropospheric ozone around Cuiabá during the dry season and smoke from biomass burning, the correlation between ozone and smoke indicates different behavior in different periods. Trajectory analyses suggest that the strong ozone peak measured in the period between August 26 and 29, 1995, may be associated not only with direct biomass‐burning emissions but also with urban/industrial emissions from big cities on the coast of Brazil and recirculation of old smoke. This view is confirmed by measurements made from the Cloud Lidar System instrument aboard the ER‐2 aircraft.

Relationships between cloud droplet effective radius, liquid water content, and droplet concentration for warm clouds in Brazil embedded in biomass smoke

During the Smoke, Clouds, and Radiation‐Brazil (SCAR‐B) project, the microphysical properties of over 1000 warm, nonprecipitating, clouds were measured from the University of Washington research aircraft. The clouds were partially embedded in the continental‐scale, smoky haze that envelops much of Brazil during the biomass‐burning season. For the entire data set, the most universal parameterization for the effective cloud droplet radius (reff) is as a function of the ratio of cloud liquid water content (LWC) to droplet concentration (essentially the volume mean radius, rv); this agrees with previous studies under less polluted conditions. Comparisons of SCAR‐B data with data from the east coast of the United States and clean oceanic areas show that the reff‐rv relationship is similar in all three cases, suggesting that even the extreme case of clouds impacted by large biomass fires can be treated similarly to more typical clouds. Beyond a certain ambient concentration of accumulation‐mode particles (∼3000–4000 cm−3), cloud drop number concentrations for cumulus clouds in Brazil were almost constant, so that further increases in the ambient particle concentration did not change reff and reff correlates well with LWC alone. For example, a cumulus cloud, which capped a particularly large smoke plume with total particle concentrations >150,000 cm−3, had the same reff‐LWC relationship as other clouds in the region where the ambient particle concentrations were ∼3000 cm−3. In this study the values of reff for cumulus clouds in Brazil affected by smoke were between 3 and 8 μm, compared to 9 to 14 μm inferred from satellite measurements of cloud reflectivity at 3.7 μm by Kaufman and Fraser [1997].

Smoke, Clouds, and Radiation‐Brazil (SCAR‐B) experiment

The Smoke, Clouds, and Radiation‐Brazil (SCAR‐B) field project took place in the Brazilian Amazon and cerrado regions in August–September 1995 as a collaboration between Brazilian and American scientists. SCAR‐B, a comprehensive experiment to study biomass burning, emphasized measurements of surface biomass, fires, smoke aerosol and trace gases, clouds, and radiation, their climatic effects, and remote sensing from aircraft and satellites. It included aircraft and ground‐based in situ measurements of smoke emission factors and the compositions, sizes, and optical properties of the smoke particles; studies of the formation of ozone; the transport and evolution of smoke; and smoke interactions with water vapor and clouds. This overview paper introduces SCAR‐B and summarizes some of the main results obtained so far. (1) Fires: measurements of the size distribution of fires, using the 50 m resolution MODIS Airborne Simulator, show that most of the fires are small (e.g., 0.005 km2), but the satellite sensors (e.g., AVHRR and MODIS with 1 km resolution) can detect fires in Brazil which are responsible for 60–85% of the burned biomass; (2) Aerosol: smoke particles emitted from fires increase their radius by as much as 60% during their first three days in the atmosphere due to condensation and coagulation, reaching a mass median radius of 0.13–0.17 μm; (3) Radiative forcing: estimates of the globally averaged direct radiative forcing due to smoke worldwide, based on the properties of smoke measured in SCAR‐B (−0.1 to −0.3 W m−2), are smaller than previously modeled due to a lower single‐scattering albedo (0.8 to 0.9), smaller scattering efficiency (3 m2g−1at 550 nm), and low humidification factor; and (4) Effect on clouds: a good relationship was found between cloud condensation nuclei and smoke volume concentrations, thus an increase in the smoke emission is expected to affect cloud properties. In SCAR‐B, new techniques were developed for deriving the absorption and refractive index of smoke from ground‐based remote sensing. Future spaceborne radiometers (e.g., MODIS on the Earth Observing System), simulated on aircraft, proved to be very useful for monitoring smoke properties, surface properties, and the impacts of smoke on radiation and climate.

Overview of atmospheric conditions during the Smoke, Clouds, and Radiation‐Brazil (SCAR‐B) field experiment

The Smoke, Clouds, and Radiation‐Brazil (SCAR‐B) field experiment was conducted in central Brazil and southern Amazonia during the period August 15 to September 20, 1995. This paper presents an overview of atmospheric conditions during the SCAR‐B period. A meteorological office was established in Brasília to support the mission flights and field activities. All the meteorological data collected during the SCAR‐B have been archived and are available to the scientific community. The meteorological conditions throughout the SCAR‐B period of August and September 1995 were near climatological conditions. Few synoptic‐scale waves traveling from southeastern Pacific reached Brazil, therefore the SCAR‐B period was characterized by long periods of low humidity, dryness, little cloudiness, or rain, and haze events. The stable situation was interrupted due to the penetration of a frontal system on September 20, 1995, and a second and more rain‐producing one on September 28, 1995. Those two frontal systems in a period of 10 days brought an end to the dry season. Air particle trajectories show that most of the smoke below 4 km from Amazonia was transported to the south and then to the east exiting the continent at 25°S–30°S. Radiative transfer calculations carried out for the clear sky gaseous and aerosol atmosphere show the decrease of the solar radiation absorption in the atmosphere‐surface system due to smoke aerosol loading at the value from 5 to 50 W m−2 depending on the aerosol optical thickness, single‐scattering albedo, and solar zenith angle.

Observations of atmospheric methane and carbon monoxide in Brazil: SCAR‐B mission

Methane (CH4) concentrations have been measured in the lower atmosphere in Brazil on board a Brazilian research aircraft. Air samples were obtained at different heights by grab sampling, and taken for analysis by flame ionization detector (FID) gas chromatography at INPE (the Brazilian Institute for Space Research). The sampling region is the cerrado area of central Brazil, around the cities of Brasilia, Cuiaba, Alta Floresta, and Porto Nacional. Instead of concentrating on plume characteristics, an effort was made to describe the average mixed atmosphere, keeping a distance from direct burning sources. The measurements of atmospheric CH4 during the dry season have shown larger mixing ratios compared to the wet season. This finding has consistently been observed for different dry seasons, but especially for the Smoke, Clouds, and Radiation: Brazil (SCAR‐B) field mission. For comparison, measurements near the ground in a region not directly affected by biomass burning (Natal, Brazil) show concentrations that are 39 parts per billion by volume (ppbv) less than the biomass burning results, which is 4 times the size of the seasonal amplitude at oceanic stations within about 10° of the equator, thus making the biomass burning values comparable to those observed at northern hemisphere midlatitudes. The vertical distribution for these regions seems to be well mixed, showing differences of about 40 ppbv between the different sampling areas of the SCAR‐B experiment. The CH4 mixing ratio was in general well correlated with the CO mixing ratio.

Measurements of irradiance attenuation and estimation of aerosol single scattering albedo for biomass burning aerosols in Amazonia

Investigation of the effects of biomass burning aerosols on the surface irradiance were conducted as a part of the Smoke, Clouds, and Radiation ‐ Brazil (SCAR‐B) experiment during August–September 1995. Measurements of broadband and spectral irradiance, in conjuction with measurements of aerosol physical and optical properties (optical depth, phase function, and size distribution) were made under varying conditions of aerosol loading during the SCAR‐B field campaign. Estimates of aerosol single scattering albedo (ω0) were made from matching of the measured irradiance values to the model computed irradiances by varying ω0, for observations made under cloudless conditions. Values of ω0, at approximately 550 nm, estimated from this technique using broadband 400–700 nm irradiance measurements, ranged from approximately 0.82 to 0.94 for the dates and times of these SCAR‐B measurements. Utilizing spectral irradiance data, the model retrieved values of ω0decreased with increasing wavelength, with the change of ω0as a function of wavelength differing on different days. Reductions in photosynthetically active radiation (PAR; 400‐700 nm) incident at the surface were computed to range from about 20 to 45% compared to background aerosol conditions for the 2 month biomass burning season at several locations in the southern Amazon Basin. These large reductions in incident PAR at the surface due to the heavy aerosol loadings could have implications for primary production of sensitive ecosystems. In addition, reductions of total incident solar radiation from aerosol direct radiative effects may have significant impact on reducing surface heating and increasing aerosol layer heating from absorption.

Remote sensing of smoke from MODIS airborne simulator during the SCAR‐B experiment

MODIS (moderate resolution imaging spectroradiometer) airborne simulator (MAS) data acquired during the SCAR‐B (Smoke, Clouds, and Radiation‐Brazil) experiment provide a test bed to evaluate aerosol retrievals for MODIS measurements over land. The SCAR‐B MAS data covered forest and cerrado regions with varying smoke concentration as a result of fires in August–September, 1995. Excellent agreement is obtained by comparing the retrieved aerosol optical thickness from the MAS data with AERONET (Aerosol Robotic Network) ground‐based Sun photometer observations. The evaluation of MODIS aerosol algorithm was performed for a range of smoke optical thickness from 0.1 to 2 at 0.66 μm wavelength over cerrado and forest sites. All the retrieved values of optical thickness are found within the anticipated retrieval errors of Δτ= ±0.05±0.2τ. This confirms for smoke aerosol the validity of the MODIS algorithms and the use of the new dynamic aerosol models in this algorithm (the algorithm was already validated previously for urban/industrial aerosol). The retrieved optical thickness is found to be very sensitive to the assumed value of the single‐scattering albedo but not sensitive to the aerosol refractive index (real part). The single‐scattering albedo of 0.90 suggested in the aerosol model yields the best results. Test of the MODIS use of a grid box of 10×10 km in the aerosol algorithm shows that the standard deviation of the retrieved aerosol should generally not be more than 10% of the average value for a large range of optical thickness values from 0.2 to 2.0, close and far from fires.

Biomass burning smoke measured using backscattered ultraviolet radiation: SCAR‐B and Brazilian smoke interannual variability

A recently developed method for monitoring biomass burning smoke from space has been applied to measurements over Brazil. Long‐term measurements of aerosol index measured from space by a series of satellites show the interannual variability of smoke‐covered areas over Brazil. Using a newly developed algorithm, the aerosol index has been inverted into aerosol optical depth values. Coincident ground‐based Sun photometer aerosol optical thickness measurements are used to validate this inversion. The interannual variability of the area covered by smoke of varying intensity is described over the period 1978 to 1995. The results show that 1995, the year of the Smoke, Clouds, and Radiation‐Brazil (SCAR‐B) mission, was similar to recent years for overall smoke area, but the area covered by heavier smoke was larger than other years in the data record.

Dynamics associated with total aboveground biomass, C, nutrient pools, and biomass burning of primary forest and pasture in Rondônia, Brazil during SCAR‐B

Burning of slashed tropical forests and pastures is among the most significant global sources of atmospheric emissions, yet the composition of the fuels and fires that creates these emissions is not well characterized. As part of the Smoke, Clouds, and Radiation‐Brazil (SCAR‐B) experiment, we measured total aboveground biomass (TAGB) as well as carbon, nitrogen, and sulfur pools in one cattle pasture and two slashed primary forests in Rondônia, Brazil. These pools were measured before and immediately after fires. From these data, we calculated the quantities of biomass and elements lost to the atmosphere during biomass burning. Prefire biomass in the pasture was 66 Mg ha−1; fire consumed 31% of this mass. Woody debris from the forest that occupied this site 12 years previously comprised 81% of the pasture prefire TAGB. Elemental inputs into the atmosphere (site losses) from the pasture fire were 9 Mg C ha−1, 88 kg N ha−1, and 5 kg S ha−1. Combining previous studies with this one, we calculate that the mean TAGB of Amazonian pastures is 74 Mg ha−1 with a mean combustion factor of 46%. Mean nutrient losses from pasture fires in Amazonia are 14 Mg C ha−1, 199 kg N ha−1, and 16 kg S ha−1. The TAGB of the two slashed primary forests before fire was 355 and 399 Mg ha−1 and following fire was 188 and 185 Mg ha−1 (i.e., a combustion factor of 47 and 54%), respectively. Combining this study with other studies of Amazon slashed primary forests, we calculate that the mean TAGB is 349 Mg ha−1 and the mean combustion factor is 48%. Total elemental losses arising from the primary forest slash fires in this study were notably higher than losses from the pasture site: 79 and 102 Mg C ha−1; 1019 and 1196 kg N ha−1; and 87 and 96 kg S ha−1. From this study combined with previous research in Rondônia and Pará, we calculate that mean nutrient losses from primary forest slash fires are 88 Mg C ha−1, 1181 kg N ha−1, and 107 kg S ha−1. As rates of deforestation are remaining high in the Brazilian Amazon and pastures are the most frequent end product, it can be expected that these will remain the dominant sources of atmospheric emissions from Amazonia in the future.

Retrieval of the real part of the refractive index of smoke particles from Sun/sky measurements during SCAR‐B

A method is used to retrieve the real part of the refractive index of ambient aerosol particles in the entire vertical column using ground‐based measurements of the angular dependence of the spectral sky radiance. The method is applied to smoke aerosol particles using spectral Sun/sky data measured by the AERONET (Aerosol Robotic Network) radiometers in Cuiabá, Brazil, during the SCAR‐B (Smoke, Clouds, and Radiation‐Brazil) experiment in 1995. The refractive index is retrieved from comparison between measurements taken in the solar almucantar and calculations using Mie theory. First the aerosol size distribution is derived from sky radiance at scattering angles less then 40°, then the refractive index is derived from sky radiances for angles of 20°–100°. Simulations and sensitivity studies are presented showing that the expected error is ±0.03, Application of the method to the Cuiabá region, which is dominated by smoke from cerrado vegetation burning, resulted in a mean value for the real part of the index of refraction of 1.53±0.04, 1.55±0.04, 1.59±0.02, and 1.58±0.01, respectively, for wavelengths of 438, 670, 870, and 1020 nm. Though we do not have independent verification of the results, we tested the effect of water vapor on the refractive index. The low humidification factors measured in Brazil and the lack of high relative humidities suggested a small effect of water vapor. In fact, as expected, a nonsignificant correlation was observed between the retrieved values of refractive index and total precipitable water vapor. Application to aerosol in the eastern United States (not reported here), with high humidity and high humidification factors, did show a strong reduction of the refractive index with increase of the total precipitable water vapor, thus generating confidence in the methodology.

Discriminating heavy aerosol, clouds, and fires during SCAR‐B: Application of airborne multispectral MAS data

A multispectral scanning spectrometer was used to obtain measurements of the reflection function and brightness temperature of smoke, clouds, and terrestrial surfaces at 50 discrete wavelengths between 0.55 and 14.2 μm. These observations were obtained from the NASA ER‐2 aircraft as part of the Smoke, Clouds, and Radiation‐Brazil (SCAR‐B) campaign, conducted over a 1500×1500 km region of cerrado and rain forest throughout Brazil between August 16 and September 11, 1995. Multispectral images of the reflection function and brightness temperature in 10 distinct bands of the MODIS airborne simulator (MAS) were used to derive a confidence in clear sky (or alternatively the probability of cloud), shadow, fire, and heavy aerosol. In addition to multispectral imagery, monostatic lidar data were obtained along the nadir ground track of the aircraft and used to assess the accuracy of the cloud mask results. This analysis shows that the cloud and aerosol mask being developed for operational use on the moderate‐resolution imaging spectroradiometer (MODIS), and tested using MAS data in Brazil, is quite capable of separating cloud, aerosol, shadow, and fires during daytime conditions over land.

SCAR‐B fires in the tropics: Properties and remote sensing from EOS‐MODIS

Two moderate resolution imaging spectroradiometer (MODIS) instruments are planned for launch in 1999 and 2000 on the NASA Earth Observing System (EOS) AM‐1 and EOS PM‐1 satellites. The MODIS instrument will sense fires with designated 3.9 and 11 μm channels that saturate at high temperatures (450 and 400 K, respectively). MODIS data will be used to detect fires, to estimate the rate of emission of radiative energy from the fire, and to estimate the fraction of biomass burned in the smoldering phase. The rate of emission of radiative energy is a measure of the rate of combustion of biomass in the fires. In the Smoke, Clouds, and Radiation‐Brazil (SCAR‐B) experiment the NASA ER‐2 aircraft flew the MODIS airborne simulator (MAS) to measure the fire thermal and mid‐IR signature with a 50 m spatial resolution. These data are used to observe the thermal properties and sizes of fires in the cerrado grassland and Amazon forests of Brazil and to simulate the performance of the MODIS 1 km resolution fire observations. Although some fires saturated the MAS 3.9 μm channel, all the fires were well within the MODIS instrument saturation levels. Analysis of MAS data over different ecosystems, shows that the fire size varied from single MAS pixels (50×50 m) to over 1 km2. The 1×1 km resolution MODIS instrument can observe only 30–40% of these fires, but the observed fires are responsible for 80 to nearly 100% of the emitted radiative energy and therefore for 80 to 100% of the rate of biomass burning in the region. The rate of emission of radiative energy from the fires correlated very well with the formation of fire burn scars (correlation coefficient = 0.97). This new remotely sensed quantity should be useful in regional estimates of biomass consumption.

Airborne spectral measurements of surface anisotropy during SCAR‐B

During the Smoke, Clouds, and Radiation‐Brazil (SCAR‐B) deployment, angular distributions of spectral reflectance for vegetated surfaces and smoke layers were measured using the scanning cloud absorption radiometer (CAR) mounted on the University of Washington C‐131A research aircraft. The CAR contains 13 narrowband spectral channels between 0.3 and 2.3 μm with a 190° scan aperture (5° before zenith to 5° past nadir) and 1° instantaneous field of view. The bidirectional reflectance is obtained by flying a clockwise circular orbit above the surface, resulting in a ground track ∼3 km in diameter within about 2 min. Although the CAR measurements are contaminated by minor atmospheric effects, results show distinct spectral characteristics for various types of surfaces. Spectral bidirectional reflectances of three simple and well‐defined surfaces are presented: cerrado (August 18, 1995) and dense forest (August 25, 1995), both measured in Brazil under nearly clear‐sky conditions, and thick smoke layers over dense forest (September 6 and 11, 1995). The bidirectional reflectances of cerrado and dense forest revealed fairly symmetric patterns along the principal plane, with varying maximal strengths and widths spectrally in the backscattering direction. In the shortwave‐infrared region the aerosol effect is very small due to low spectral optical depth. Also, these backscattering maxima can be seen on the bidirectional reflectance of smoke layer over dense forest. These detailed measurements of the angular distribution of spectral reflectance can be parameterized by a few independent variables and utilized to retrieve either surface characteristics or aerosol microphysical and optical properties (e.g., size distribution and single‐scattering parameters), if proper physical and radiation models are used. The spectral‐hemispherical albedo of these surfaces is obtained directly by integrating all angular measurements and is compared with the measured nadir reflectance. Using CAR nadir reflectance as a surrogate for spectral‐hemispherical albedo can cause albedos to be underestimated by 10–60%, depending on solar zenith angle.

Sphericity and morphology of smoke particles from biomass burning in Brazil

The degree of nonsphericity of smoke particles from biomass burning in Brazil was measured aboard the University of Washington C‐131A aircraft during the Smoke, Clouds, and Radiation‐Brazil (SCAR‐B) project for several regions, types of fuel, and combustion. The nonsphericity (αo) of the particles was obtained from electrooptical light‐scattering measurements, using an aerosol asymmetry analyzer, and from scanning electron microscopy (SEM) images of the particles. The electrooptical measurements provide a measure of the nonsphericity based on the difference between the light scattering coefficient for aligned and randomly oriented particles. The SEM photographs provide information on the geometric shapes of the particles. The maximum value of αo obtained during the SCAR‐B for biomass burning in Brazil was below 13%. The degree of nonsphericity of the particles is shown to be related to the combustion efficiency, the mass absorption efficiency, and the fraction of black carbon to total particle mass. It is concluded after smoke particles from biomass burning in Brazil have been in the atmosphere for more than about 1 hour that the spherical approximation (and therefore Mie theory) is reasonably valid for estimating the physical and optical properties of the particles.