A per-pixel, non-stationary mixed model for empirical line atmospheric correction in remote sensing

Abstract Atmospheric correction is a key stage in the processing of remotely sensed data. The empirical line method (ELM) is used widely to correct at-sensor radiance or DN to at-surface reflectance. It is based on a simple linear relationship between those two variables. Effective application of the model requires that it is estimated in a precise and unbiased fashion. The usual approach is to use ordinary least squares (OLS) regression to model the relationship between the average reflectance and radiance for a small number (3 to 8) of ground targets (GTs) and then to apply the regression on a per-pixel basis to the image. This leads to a mismatch between the scale at which the model is estimated and the scale at which the model is applied. Further, this approach wastes information and can lead to inconsistent estimators. These problems are addressed in the new approach presented here. The model was estimated on a per-pixel rather than per-GT basis. This yielded consistent, precise estimators for the ELM, but placed stronger requirements on the modeling. Specifically spatial autocorrelation and non-constant variance (heteroskedasticity) in the model residuals needed to be addressed. This was undertaken using the linear mixed model (LMM), which is a model-based expression of the geostatistical method. Of particular interest is the use of a non-stationary LMM to address the heteroskedasticity. The approach taken in this paper is of significance for a broader set of remote sensing applications. Regression and geostatistics are often applied, based typically on a stationary model. This paper shows how heteroskedasticity can be assessed and modeled using the non-stationary LMM. Heteroskedasticity is present in other remote sensing applications hence the non-stationary modeling approach, demonstrated here, is likely to be beneficial.

[1]  Michael D. Steven,et al.  Estimation of atmospheric corrections from multiple aircraft imagery , 1986 .

[2]  Xingfa Gu,et al.  Evaluation of measurement errors in ground surface reflectance for satellite calibration , 1992 .

[3]  M. Kenward,et al.  Small sample inference for fixed effects from restricted maximum likelihood. , 1997, Biometrics.

[4]  Michael L. Stein,et al.  Local likelihood estimation for nonstationary random fields , 2011, J. Multivar. Anal..

[5]  Carol A. Gotway,et al.  Statistical Methods for Spatial Data Analysis , 2004 .

[6]  Edward J. Milton,et al.  A portable multiband radiometer for ground data collection in remote sensing , 1980 .

[7]  Edzer J. Pebesma.  The Role of External Variables and GIS Databases in Geostatistical Analysis , 2006, Trans. GIS.

[8]  D. P. Groeneveld,et al.  Empirical proof of the empirical line , 2008 .

[9]  David Higdon,et al.  A process-convolution approach to modelling temperatures in the North Atlantic Ocean , 1998, Environmental and Ecological Statistics.

[10]  R. Lark,et al.  A linear mixed model, with non-stationary mean and covariance, for soil potassium based on gamma radiometry , 2010 .

[11]  J. Townshend,et al.  An operational atmospheric correction algorithm for Landsat Thematic Mapper imagery over the land , 1997 .

[12]  R. Lark,et al.  Model‐based analysis using REML for inference from systematically sampled data on soil , 2004 .

[13]  R. Lark,et al.  On spatial prediction of soil properties in the presence of a spatial trend: the empirical best linear unbiased predictor (E‐BLUP) with REML , 2006 .

[14]  C. Woodcock,et al.  Classification and Change Detection Using Landsat TM Data: When and How to Correct Atmospheric Effects? , 2001 .

[15]  Richard Webster,et al.  Regression and functional relations , 1997 .

[16]  R. M. Lark,et al.  Modelling non-stationary variance of soil properties by tempering an empirical spectrum , 2009 .

[17]  A. Goetz,et al.  Atmospheric correction algorithms for hyperspectral remote sensing data of land and ocean , 2009 .

[18]  T. Malthus,et al.  The empirical line method for the atmospheric correction of IKONOS imagery , 2003 .

[19]  S. Liang.  Quantitative Remote Sensing of Land Surfaces , 2003 .

[20]  John R. Schott,et al.  Remote Sensing: The Image Chain Approach , 1996 .

[21]  Juha Suomalainen,et al.  The selection of appropriate spectrally bright pseudo-invariant ground targets for use in empirical line calibration of SPOT satellite imagery , 2011 .

[22]  Hongliang Fang,et al.  Atmospheric correction of Landsat ETM+ land surface imagery. I. Methods , 2001, IEEE Trans. Geosci. Remote. Sens..

[23]  Edward J. Milton,et al.  On the temporal stability of ground calibration targets: implications for the reproducibility of remote sensing methodologies , 2006 .

[24]  Christopher J Paciorek,et al.  Spatial modelling using a new class of nonstationary covariance functions , 2006, Environmetrics.

[25]  Edward J. Milton,et al.  Review Article Principles of field spectroscopy , 1987 .

[26]  D. Bates,et al.  Mixed-Effects Models in S and S-PLUS , 2001 .

[27]  Hongliang Fang,et al.  Atmospheric correction of Landsat ETM+ land surface imagery: Part I: Methods , 2001 .

[28]  Edzer Pebesma,et al.  Spatial aggregation and soil process modelling , 1999 .

[29]  P. Chavez.  An improved dark-object subtraction technique for atmospheric scattering correction of multispectral data , 1988 .

[30]  Diofantos G. Hadjimitsis,et al.  The use of selected pseudo-invariant targets for the application of atmospheric correction in multi-temporal studies using satellite remotely sensed imagery , 2009, Int. J. Appl. Earth Obs. Geoinformation.

[31]  Emmanuelle Vaudour,et al.  Spatial retrieval of soil reflectance from SPOT multispectral data using the empirical line method , 2008 .

[32]  Y.‐Q. Jin,et al.  Suspended sediment concentrations in the Yangtze River estuary retrieved from the CMODIS data , 2006 .

[33]  Michael H. Kutner.  Applied Linear Statistical Models , 1974 .

[34]  R. M. Lark,et al.  Kriging a soil variable with a simple nonstationary variance model , 2009 .

[35]  Gordon K. Smyth,et al.  An Efficient Algorithm for REML in Heteroscedastic Regression , 2002 .

[36]  Alfred Stein.  Some basic elements of statistics , 1999 .

[37]  Kurtis J. Thome,et al.  A refined empirical line approach for reflectance factor retrieval from Landsat-5 TM and Landsat-7 ETM+ , 2001 .

[38]  Peter M. Atkinson,et al.  On the effect of positional uncertainty in field measurements on the atmospheric correction of remotely sensed imagery , 2004 .

[39]  K. Anderson,et al.  Sources of uncertainty in vicarious calibration: understanding calibration target re .ectance , 2003, IGARSS 2003. 2003 IEEE International Geoscience and Remote Sensing Symposium. Proceedings (IEEE Cat. No.03CH37477).

[40]  E. Milton,et al.  The use of the empirical line method to calibrate remotely sensed data to reflectance , 1999 .

[41]  P. Diggle,et al.  Model‐based geostatistics , 2007 .

[42]  M. Fuentes.  Spectral methods for nonstationary spatial processes , 2002 .