Yoram J. Kaufman

Algorithm for automatic atmospheric corrections to visible and near-IR satellite imagery

Abstract An algorithm is developed for automatic atmospheric correction of satellite imagery of the Earths surface. The algorithm is based solely on the satellite image being corrected and on climatology of the area. It is applicable to low resolution (1 km field of view) and high resolution (10-80m field of view) imagery of land areas for the solar spectrum. The algorithm requires that some pixels in the image will correspond to dense dark vegetation as the surface cover. Once the presence of such pixels is established, the algorithm automatically chooses these pixels, derives the atmospheric optical thickness (a measure of the amount of haze) and corrects the image. The algorithm is sensitive to the assumed reflectance of the dense dark vegetation. As a result, the accuracy of the corrected surface reflectance (p) is expected to be δp-±0.01. It is not very sensitive to the assumed aerosol characteristics, the accuracy of satellite calibration or the knowledge of the exact fraction of the image covered ...

The Relative Importance of Aerosol Scattering and Absorption in Remote Sensing

Previous attempts to explain the effect of aerosols on satellite measurements of surface properties for the visible and near-infrared spectrum have emphasized the amount of aerosols without consideration of their absorption properties. In order to estimate the importance of absorption, the radiances of the sunlight scattered from models of the Earth-atmosphere system are computed as functions of the aerosol optical thickness and absorption. The absorption effect is small where the surface reflectance is weak, but is important for strong reflectance. These effects on classification of surface features, measuring vegetation index, and measuring surface reflectance are presented.

Size distribution and scattering phase function of aerosol particles retrieved from sky brightness measurements

Ground-based measurements of the solar transmission and sky radiance in a horizontal plane through the Sun are taken in several geographical regions and aerosol types: dust in a desert transition zone in Israel, sulfate particles in Eastern and Western Europe, tropical aerosol in Brazil, and mixed continental/maritime aerosol in California. Stratospheric aerosol was introduced after the eruption of Mount Pinatubo in June 1991. Therefore measurements taken before the eruption are used to analyze the properties of tropospheric aerosol; measurements from 1992 are also used to detect the particle size and concentration of stratospheric aerosol. The measurements are used to retrieve the size distribution and the scattering phase function at large scattering angles of the undisturbed aerosol particles. The retrieved properties represent an average on the entire atmospheric column. A comparison between the retrieved phase function for a scattering angle of 120°, with phase function predicted from the retrieved size distribution, is used to test the assumption of particle homogeneity and sphericity in radiative transfer models (Mie theory). The effect was found to be small (20%±15%). For the stratospheric aerosol (sulfates), as expected, the phase function was very well predicted using the Mie theory. A model with a power law size distribution, based on the spectral dependence of the optical thickness, a, cannot estimate accurately the phase function (up to 50% error for λ = 0.87 μm). Before the Pinatubo eruption the ratio between the volumes of sulfate and coarse particles was very well correlated with α. The Pinatubo stratospheric aerosol destroyed this correlation. The aerosol optical properties are compared with analysis of the size, shape, and composition of the individual particles by electron microscopy of in situ samples. The measured volume size distributions before the injection of stratospheric aerosol consistently show two modes, sulfate particles with rm 0.7 μm. The “window” in the tropospheric aerosol in this radius range was used to observe a stable stratospheric aerosol in 1992, with rm ∼ 0.5 μm. A combination of such optical thickness and sky measurements can be used to assess the direct forcing and the climatic impact of aerosol. Systematic inversion for the key aerosol types (sulfates, smoke, dust, and maritime aerosol) of the size distribution and phase function can give the relationship between the aerosol physical and optical properties that can be used to compute the radiative forcing. This forcing can be validated in dedicated field experiments.

Atmospheric Effects On Remote Sensing Of Surface Reflectance

This paper reviews the atmospheric effects on remote sensing of surface reflectance. The scattering and absorption of sunlight by atmospheric molecules and aerosols affects the quality of images of the surface remotely sensed from satellites and aircrafts. The concentration and characteristics of the atmospheric aerosols vary from place to place and vary with time. The effect of atmospheric aerosols on the upward radiance depends on their optical thickness, scattering phase function and absorption. These parameters result from the aerosol concentration, composition, and the relative humidity. For high resolution images the aerosol scale height is also of importance. The radiative transfer theory that predicts the atmospheric radiances for a given surface and atmosphere is a well established theory for the case of uniform surfaces (or low resolution data). Some radiative transfer models exist for nonuniform surfaces and others are being developed. Recent field experiment and laboratory simulation data confirm the need for these models and can be used for their testing. It is shown that the atmospheric effect reduces the apparent resolution of satellite imagery and causes errors in the classification of surface fields. Suggestions for correction procedures are given. Such corrections can be based on ground observations, on satellite radiances above dark areas, or on climatologic information, depending on the accuracy of the corrections needed. The chosen correction algorithm depends also on the image resolution and the specific remote sensing application.

Fast algorithms for removing atmospheric effects from satellite images

The varied features of the earths surface each reflect sunlight and other wavelengths of solar radiation in a highly specific way. This principle provides the foundation for the science of satellite based remote sensing. A vexing problem confronting remote sensing researchers, however, is that the reflected radiation observed from remote locations is significantly contaminated by atmospheric particles. These aerosols and molecules scatter and absorb the solar photons reflected by the surface in such a way that only part of the surface radiation can be detected by a sensor. The article discusses the removal of atmospheric effects due to scattering and absorption, ie., atmospheric correction. Atmospheric correction algorithms basically consist of two major steps. First, the optical characteristics of the atmosphere are estimated. Various quantities related to the atmospheric correction can then be computed by radiative transfer algorithms, given the atmospheric optical properties. Second, the remotely sensed imagery is corrected by inversion procedures that derive the surface reflectance. We focus on the second step, describing our work on improving the computational efficiency of the existing atmospheric correction algorithms. We discuss a known atmospheric correction algorithm and then introduce a substantially more efficient version which we have devised. We have also developed a parallel implementation of our algorithm.

Remote sensing of biomass burning in the tropics

Abstract A new method is developed for the global assessment of the contribution of biomass burning to climate change (trace gases and particulates emission). The method is based directly on remote sensing of the emitted products - particulates. The detected mass of emitted particulates is converted into the mass of the emitted trace gases using published relations between the particulates and trace gases emitted, for the flaming and smoldering phases. The NOAA-AVHRR 1km resolution images, in the visible, near IR, mid-IR and IR channels are used for this purpose. The analysis can be applied to regions where intensive biomass burning takes place. Preliminary analysis of the 1987 burning season shows that in Brazil (between 6.5°–15.5° south and 55°–67° west), during the three months of the dry season, there may be up to 5000 fires a day, observed from space, each contributing 200 ton/hr of CO 2 , 20 ton/hr of CO and 0.5 ton/hr of CH 4 to the atmosphere. During the dry season of 1987, it is estimated that 150,000 fires were burning, resulting in the emission of 6·10 12 g of particulates, 1.6·10 12 g of CH 4 , 10 14 g of Co and 7·10 14 g of CO 2 . A comparison to estimates of global emission are given.

Efficient Algorithms for Atmospheric Correction of Remotely Sensed Data

Remotely sensed imagery has been used for developing and validating various studies regarding land cover dynamics. However, the large amounts of imagery collected by the satellites are largely contaminated by the effects of atmospheric particles. The objective of atmospheric correction is to retrieve the surface reflectance from remotely sensed imagery by removing the atmospheric effects. We introduce a number of computational techniques that lead to a substantial speedup of an atmospheric correction algorithm based on using look-up tables. Excluding I/O time, the previous known implementation processes one pixel at a time and requires about 2.63 seconds per pixel on a SPARC-10 machine, while our implementation is based on processing the whole image and takes about 4-20 microseconds per pixel on the same machine. We also develop a parallel version of our algorithm that is scalable in terms of both computation and I/O. Experimental results obtained show that a Thematic Mapper (TM) image (36 MB per band, 5 bands need to be corrected) can be handled in less than 4.3 minutes on a 32-node CM-5 machine, including I/O time.

Remote sensing of biomass burning associated with deforestation

A new method has been developed for the global assessment of trace gases and particulates emission from tropical biomass burning. The method is based on remote sensing of one emitted product - particulates. It uses daily meteorological satellite data with resolution of one km2. The visible (0.63 rim) and near Infrared (0.84 rim) bands are used to determine the mass of particulates in the emitted smoke and to estimate the relative contribution of flaming and smoldering fires to the resulting smoke. The mid-IR (3.5-3.9 rim) and the thermal infrared (1 0.5-1 1 .4 ji.m) bands are used to detect and count fires in order to integrate the smoke result with the whole season and for the whole area of interest. The thermal channels are sensitive enough to detect flaming fires as small as lOm*lOm and smoldering fires as small as 30m*30m. The detected mass of emitted particulates is converted into a mass of emitted trace gases using published relations between the emitted particulates and trace gases for the flaming and smoldering phases. The.technique can be applied to regions where intensive biomass burning takes place. It is capable of monitoring the extent of current biomass burning, to discover new deforestation frontiers (unknown otherwise), and to estimate quantitative contribution of biomass burning to changes in atmospheric composition. The method has been applied to a limited area where substantial deforestation has taken place. Analysis of the 1987 burning season shows that in Brazil (in a limited area between 6.5-15.5 south and 55-67 west) during the three months of the dry season (July 1 till Sept. 30) there are up to 8000 fires a day (observed from space) each contributing 4,500 tons of C02, 750 tons of CO and 26 tons of CH4 to the atmosphere. During the dry season of 1 987, it is estimated that 240,000 fires were burning in this area resulting in the emission of 1.1013 of particulates, 71 012 g of CH4, 21 014 g of CO and 1 •1 015 g of CO2. A comparison to estimates of global emissions is also given.