Abstract
Remotely sensed land cover maps are increasingly used as inputs into environmental simulation models whose outputs inform decisions and policy-making. Risks associated with these decisions are dependent on model output uncertainty, which is in turn affected by the uncertainty of land cover inputs. This article presents a method of quantifying the uncertainty that results from potential mis-classification in remotely sensed land cover maps. In addition to quantifying uncertainty in the classification of individual pixels in the map, we also address the important case where land cover maps have been upscaled to a coarser grid to suit the users’ needs and are reported as proportions of land cover type. The approach is Bayesian and incorporates several layers of modelling but is straightforward to implement. First, we incorporate data in the confusion matrix derived from an independent field survey, and discuss the appropriate way to model such data. Second, we account for spatial correlation in the true land cover map, using the remotely sensed map as a prior. Third, spatial correlation in the mis-classification characteristics is induced by modelling their variance. The result is that we are able to simulate posterior means and variances for individual sites and the entire map using a simple Monte Carlo algorithm. The method is applied to the Land Cover Map 2000 for the region of England and Wales, a map used as an input into a current dynamic carbon flux model.
Similar content being viewed by others
References
Bayarri MJ, Berger JO, Cafeo J, Garcia-Donato G, Liu F, Palomo J, Parthasarathy RJ, Paulo R, Sacks J, Walsh D (2007) Computer model validation with functional output. Ann Stat 35(5):1874–1906
Besag JE (1986) On the statistical analysis of dirty pictures. J R Stat Soc B 48:259–302
Conti S, O’Hagan A (2010) Bayesian emulation of complex multi-output and dynamic computer models. J Stat Plan Inference 140(3):640–651
De Bruin S (2000) Predicting the areal extent of land-cover types using classified imagery and geostatistics. Remote Sens Environ 74:387–396
Denham R, Mengersen K, Witte C (2009) Bayesian analysis of thematic map accuracy data. Remote Sens Environ 113:371–379
Divino F, Frigessi A, Green PJ (2000) Penalized pseudolikelihood inference in spatial interaction models with covariates. Scand J Stat 27:445–458
Eroğlu H, Çakir G, Sivrikaya F, Akay AE (2010) Using high resolution images and elevation data in classifying erosion risks of bare soil areas in the Hatila Valley Natural Protected Area, Turkey. Stoch Environ Res Risk Assess 24:699–704
Foody GM (2002) Status of land cover classification accuracy assessment. Remote Sens Environ 80:185–201
Foody GM (2007) The evaluation and comparison of thematic maps derived from remote sensing. In: 7th international symposium on spatial accuracy assessment and environmental sciences, Lisbon, pp 18–31
Frigessi A, Stander J (1994) Informative priors for the Bayesian classification of satellite images. J Am Stat Assoc 89(426):703–709
Fuller RM, Smith GM, Sanderson JM, Hill RA, Thomson AG (2002) Countryside survey 2000, Module 7. Land Cover Map 2000. Final Report, Technical Report
Fuller RM, Cox R, Clarke RT, Rother P, Hill RA, Smith GM, Thomson AG, Brown NJ, Howard DC, Stott AP (2005) The UK Land Cover Map 2000: planning, construction and calibration of a remotely sensed, user-orientated map of broad habitats. Int J Appl Earth Obs Geoinf 7:202–216
Geman S, Geman D (1984) Stochastic relation, the Gibbs distribution and the Bayesian restoration of images. IEEE Trans Pattern Anal Mach Intell 6(6):721–741
Giri C, Zhu Z, Reed B (2005) A comparative analysis of the Global Land Cover 2000 and MODIS land cover data sets. Remote Sens Environ 94:123–132
Green EJ, Strawderman WE (1994) Determining the accuracy of thematic maps. The Statistician 43(1):77–85
Hansen M, Reed B (2000) A comparison of the IGBP DISCover and University of Maryland 1 km global land cover products. Int J Remote Sens 21:1365–1373
Houghton JT, Ding Y, Griggs DJ, Nouger M, van der Linden PJ, Dai X, Maskell K, Johnson CA (2001) IPCC(2001) climate change 2001: the scientific basis. Contribution of Working Group 1 to the Third Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, IPCC
Huang Q, Cai Y (2007) Simulation of land use change using GIS-based stochastic model: the case study of Shiqian County, Southwestern China. Stoch Environ Res Risk Assess 21:419–426
Hwang Y, Clark MP, Rajagopalan B (2011) Use of daily precipitation uncertainties in streamflow simulation and forecast. Stoch Environ Res Risk Assess 25:957–972
Intergovernmental Panel on Climate Change (2007) Climate change 2007: Synthesis Report
Jung M, Henkel K, Herold M, Churkina G (2006) Exploiting synergies of global land cover products for carbon cycle modelling. Remote Sens Environ 101(4):534–553
Kennedy M, O’Hagan A, Anderson CW, Lomas M, Woodward FI, Heinemeyer A, Gosling JP (2008) Quantifying uncertainty in the biospheric carbon flux for England and Wales. J R Stat Soc A 171(1):109–135
Klein R, Press SJ (1989) Contextual Bayesian classification of remotely sensed data. Commun Stat A 18:3177–3202
Klein R, Press SJ (1990a) Bayesian classification of remotely sensed data when training data is part of the scene. Rev Bras Probab Estat 4:43–67
Klein R, Press SJ (1990b) Bayesian contextual classification with neighbors correlated with training data, Bayesian and likelihood methods in statistics and econometrics: essays in Honor of George A. Bernard, pp 337–355
Kriauciuniene J, Jakimavicius D, Sarauskien D, Kaliatka T (2012) Estimation of uncertainty sources in the projections of Lithuanian river runoff. Stoch Environ Res Risk Assess. doi:10.1007/s00477-012-0608-7
Li W, Zhang C (2011) Land-cover classification with uncertainty assessment based on expert-interpreted pixels from remotely sensed imagery. IEEE Trans Geosci Remote Sens 49(8):2983–2992
Maselli F, Conese C, Petkov L (1994) Use of probability entropy for the estimation and graphical representation of the accuracy of maximum likelihood classifications. ISPRS J Photogramm Remote Sens 49(2):13–20
Mayaux P, Eva H, Gallego J, Strahler AH, Herold M, Agrawal S, Naumov S, De Miranda EE, Di Bella CM, Ordoyne C, Kopin Y, Roy P (2006) Validation of the Global Land Cover 2000 map. IEEE Trans Geosci Remote Sens 44(7):1728–1739
McCallum I, Obersteiner M, Nilsson S, Shivdenko A (2006) A spatial comparison of four satellite derived 1 km global land cover datasets. Appl Earth Obs Geoinf 8:246–255
O’Hagan A (2006) Bayesian analysis of computer code outputs: a tutorial. Reliab Eng Syst Saf 91:1290–1300
Quaife T, Quegan S, Disney M, Lewis P, Lomas M, Woodward FI (2008) The impact of errors in satellite-derived land cover on estimates of biosphere carbon fluxes. Glob Biogeochem Cycles 22:GB4016
Sacks J, Welsh WJ, Mitchell TJ, Wynn HP (1989) Design and analysis of computer experiments. Stat Sci 4:409–423
Schneider S, Lane J (2005) Integrated assessment modeling of global climate change: much has been learned—still a long and bumpy road ahead. Integr Assess J 5(1):41–75
Sivakumar B, Christako G (2011) Climate: patterns, changes and impacts. Stoch Environ Res Risk Assess 25:443–444
Stehman SV, Wickham JD, Smith JH, Yang L (2003) Accuracy of the 1992 National Land-Cover Data for the eastern United States: statistical methodology and regional results. Remote Sens Environ 86:500–516
Storvik G, Fjortoft R, Solberg AA (2005) A Bayesian approach to classification of multiresolution remote sensing data. IEEE Trans Geosci Remote Sens 43(3):539–547
Switzer P (1980) Extensions of linear discriminant analysis for statistical classification of remotely sensed satellite imagery. Math Geol 12:367–376
Wang J, Haining R, Zhidong C (2010) Sample surveying to estimate the mean of a heterogeneous surface: reducing the error variance through zoning. Int J Geogr Inf Sci 24(4):523–543
Weidong L, Zhang C, Dey DK (2010) Estimating threshold-exceeding probability maps of environmental variables with Markov chain random fields. Stoch Environ Res Risk Assess 24:1113–1126
Willems P (2001) Stochastic description of the rainfall input errors in lumped hydrological models. Stoch Environ Res Risk Assess 15:132–152
Woodward FI, Lomas MR (2004) Vegetation dynamics—simulating responses to climatic change. Biol Rev 79(3):643–670
Wu DD (2011) Introduction to the special SERRA issue on “risks, uncertainties and the environment”. Stoch Environ Res Risk Assess 25:301–304
Acknowledgments
This project was funded by the NERC Centre for Terrestrial Carbon Dynamics and the NERC National Centre for Earth Observation.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Cripps, E., O’Hagan, A. & Quaife, T. Quantifying uncertainty in remotely sensed land cover maps. Stoch Environ Res Risk Assess 27, 1239–1251 (2013). https://doi.org/10.1007/s00477-012-0660-3
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00477-012-0660-3