A VNIR/SWIR atmospheric correction algorithm for hyperspectral imagery with adjacency effect Lee C. Sanders a, *, John R. Schott b , Rolando Raquen ˜o b a Veridian-ERIM International, PO Box 134008, Ann Arbor, MI 48113 4008, USA b Center for Imaging Science, Rochester Institute of Technology, Rochester, NY, USA Received 16 October 2000; accepted 5 March 2001 Abstract Radiometrically calibrated hyperspectral imagery contains information relating to the material properties of a surface target and the atmospheric layers between the surface target and the sensor. All atmospheric layers contain well-mixed molecular gases, aerosol particles, and water vapor, and information about these constituents may be extracted from hyperspectral imagery by using specially designed algorithms. This research describes a total sensor radiance-to-ground reflectance inversion program. An equivalent surface-pressure depth can be extracted using the Non-Linear Least-Squares Spectral Fit (NLLSSF) technique on the 760-nm oxygen band. Two different methods, the Atmospheric Pre-Corrected Differential Absorption (APDA) and NLLSSF, can be used to derive total columnar water vapor using the radiative transfer model MODTRAN 4.0. Atmospheric visibility can be derived via the NLLSSF technique from the 400 – 700-nm bands or using an approach that uses the upwelled radiance fit from the Regression Intersection Method from 550 to 700 nm. A new numerical approximation technique is also introduced to calculate the effect of the target surround on the sensor-received radiance. The recovered spectral reflectances for each technique are compared to reflectance panels with well-characterized ground truth. D 2001 Elsevier Science Inc. All rights reserved. Keywords: Atmospheric correction; Hyperspectral; Inversion; Reflectance; NLLSSF; APDA; Visibility; Aerosols; Scattering phase function 1. Introduction For many years, the astronomical community has used spectroscopy to determine the chemical composition of stellar objects. The atomic and molecular constituents of stars, planets, and nebulae have been revealed by their unique spectra that, in turn, are due to their different properties of absorption and emission of electromagnetic energy. The spectral signatures of these elements arise from their electronic, vibrational, and rotational transitions. Sim- ilar information is being extracted from air- and space-borne instruments to access properties about the earth’s surface structure and the composition of the atmosphere. It would be useful if the controlled approach of spectro- scopy could be applied to airborne or space-based imaging spectrometry of the earth. The calculus of atmospheric characterization and identification of the constituents of ground objects would be simplified. Unfortunately, this calculation is not trivial. The earth’s atmosphere is a com- plex mix of molecular and larger sized compounds that are in flux spatially and temporally. To determine the scene content of an image with confidence, the atmosphere must be characterized to sufficient accuracy to obtain ground reflectance units to one reflectance unit or to estimate temperature parameters to a half kelvin. To date, the best methods for extracting atmospheric information rely heavily on the combination of ground truth measurements of targets in the scene and ground-based atmospheric measurements (for aerosol and water vapor determination) made with sun-photometers and radiosondes. Obtaining ground truth is an expensive, laborious, and time- consuming task. For physical and economic reasons, few multispectral or imaging spectrometer missions measure actual conditions simultaneously with image acquisition. To fill this computational void, algorithms have been developed to extract atmospheric data directly from the spectra of individual pixels in the hyperspectral image. All such algorithms use some form of radiative transfer model of the atmosphere. 0034-4257/01/$ – see front matter D 2001 Elsevier Science Inc. All rights reserved. PII:S0034-4257(01)00219-X * Corresponding author. Tel.: +1-734-994-1200/+1-734-994-2293; fax: +1-734-994-5704. E-mail address: sanders@erim-int.com (L.C. Sanders). www.elsevier.com/locate/rse Remote Sensing of Environment 78 (2001) 252 – 263