Spatiotemporal analysis of wind speed via the Bayesian maximum entropy approach

  • Özlem BaydaroğluEmail author
  • Kasım Koçak
Original Article


Spatiotemporal mapping of wind speed data is of great importance on determining the wind energy potential, construction zones, projecting future values and planning resources. From this point of view, monthly and yearly mean wind speed maps of Turkey are generated using the Bayesian maximum entropy (BME) approach. The BME, a nonlinear geostatistical approach, is the only method which uses not only data on the process but also all information about the data in a spatiotemporal mapping. During processing data, physical laws, hypotheses, experiences, scientific theories, high order space/time moments, various types of uncertain information, outputs of models, etc. are incorporated to the process. Hence, it maximizes knowledge about the variable processed because it uses exact and auxiliary data which belong to the regarding variable. In this study, Turkey daily mean wind speed data (m/s) measured at 10 m between the years 2010 and 2015 have been used. 0.05°, 0.1°, 0.3°, 0.5° and 1.0° have been chosen for spatial ranges and 1 day, week, month have been used for temporal ranges to determine the most appropriate intervals in the yearly forecast after lots of trials. Different kinds of local means are used so as to find proper kriging method. In the forecast process, hard data (raw data), hard data with soft data (auxiliary data), detrended hard with soft data are implemented, separately. All studies show that using hard with soft data gives the best forecast results and the most appropriate kriging method is determined as the ordinary kriging. Similarly to the other techniques which involve kriging methods, variances of errors are taken as a performance criterion. When the soft data are used, variances of the forecast errors reduce by half. In addition, all forecast results stay the interval of confidence levels and error variances of forecast results are quite low. Moreover, the BME forecast is employed for each month. Differently from the yearly forecast, the application of the detrended hard and soft data is found the most convenient experiment because it decreases the error values. From this point forth, it can be said that the usage of the soft data increases the accuracy of the forecasts. In addition, unknown wind speed values due to the lack of the stations are calculated and even wind speed values over the seas are identified. The study is the first and only spatiotemporal wind speed study used the BME approach for Turkey. It is thought that the mean wind speed map for Turkey can be considered as a current wind atlas of Turkey.


Bayesian maximum entropy (BME) Spatiotemporal analysis Forecast Wind speed Wind atlas 



The study is a part of Ph.D. thesis which has won the award of the best Ph.D. thesis in Istanbul Technical University in 2016.


  1. Adam-Poupart A, Brand A, Fournier M, Jerrett M, Smargiassi A (2014) Spatiotemporal modeling of ozone levels in Quebec (Canada): a comparison of kriging, land-use regression (LUR), and combined Bayesian maximum entropy-LUR approaches. Environ Health Perspect 122(9):970–976CrossRefGoogle Scholar
  2. Akita Y, Chen JC, Serre ML (2012) The moving-window Bayesian maximum entropy framework: estimation of PM2.5 yearly average concentration across the contiguous United States. J Expos Sci Environ Epidemiol 22(5):496–501CrossRefGoogle Scholar
  3. Akita Y, Baldasano JM, Beelen R, Cirach M, De Hoogh K, Hoek G, Nieuwenhuijsen M, Serre ML, De Nazelle A (2014) Large scale air pollution estimation method combining land use regression and chemical transport modeling in a geostatistical framework. Environ Sci Technol 48(8):4452–4459CrossRefGoogle Scholar
  4. Bayat B, Nasseri M, Gholamreza N (2014) Improving Bayesian maximum entropy and ordinary Kriging methods for estimating precipitations in a large watershed: a new cluster-based approach. Canadian J Earth Sci 51(1):43–55CrossRefGoogle Scholar
  5. Bayesian Maximum Entropy Graphical Users Interface (BMEGUI) (2014) University of North Carolina. For the BMEGUI version 3.0.1. Accessed 1 May 2015
  6. Bogaert P, Christakos G (1997) Spatiotemporal analysis and processing of thermometric data over Belgium. J Geophys Res Atmos 102:25831–25846CrossRefGoogle Scholar
  7. Choi KM, Christakos G, Wilson ML (2006) El Nino effects on influenza mortality risks in the state of California. Public Health 120(6):505–516CrossRefGoogle Scholar
  8. Christakos G (1990) A Bayesian maximum entropy view to the spatial estimation problem. Math Geol 22(7):763–776CrossRefGoogle Scholar
  9. Christakos G (1991a) Some applications of the Bayesian, maximum-entropy concept in geostatistics. In: Maximum entropy and Bayesian methods, Fundamental theories of physics. Kluwer Academic Publishers, Dordrecht, pp 215–229CrossRefGoogle Scholar
  10. Christakos G (1991b) A theory of spatiotemporal random fields and its application to space–time data processing. IEEE Trans Syst Man Cybern 21(4):861–875CrossRefGoogle Scholar
  11. Christakos G (1992) Random field models in earth sciences. Academic Press, San DiegoGoogle Scholar
  12. Christakos G (1998a) Spatiotemporal information systems in soil and environmental sciences. Geoderma 85(2):141–179CrossRefGoogle Scholar
  13. Christakos G (1998b) Multi-point BME space/time mapping of environmental variables. In: Burganos VN, Karatzas GP, Payatakes AC, Gray WG, Pinder GF (eds) Computational methods in water resources XII. Computational methods in surface and groundwater transport, vol 2. Computational Mechanics Publications, Southampton, pp 289–296Google Scholar
  14. Christakos G (2000) Modern spatiotemporal geostatistics. Oxford University Press, OxfordGoogle Scholar
  15. Christakos G (2008) Bayesian maximum entropy. In: Kanevski M (ed) Advanced mapping of environmental data, geostatistics, machine learning and Bayesian maximum entropy. Geographical information systems series. Wiley, Hoboken, NJ, pp 247–306Google Scholar
  16. Christakos G, Kolovos A, Serre ML, Vukovich F (2004) Total ozone mapping by integration databases from remote sensing instruments and empirical models. IEEE Trans Geosci Remote Sens 42(5):991–1008CrossRefGoogle Scholar
  17. De Nazelle A, Arunachalam S, Serre ML (2010) Bayesian maximum entropy integration of ozone observations and model predictions: an application for attainment demonstration in North Carolina. Environ Sci Technol 44(15):5707–5713CrossRefGoogle Scholar
  18. Douaik A, Van Meirvenne M, Toth T, Serre M (2004) Space–time mapping of soil salinity using probabilistic bayesian maximum entropy. Stoch Environ Res Risk Assess 18(4):219–227CrossRefGoogle Scholar
  19. Fan L, Xiao Q, Wen J, Liu Q, Jin R, You D, Li X (2015) Mapping high-resolution soil moisture over heterogeneous cropland using multi-resource remote sensing and ground observations. Remote Sens 7(10):13273–13297CrossRefGoogle Scholar
  20. Gao S, Zhu Z, Liu S, Jin R, Yang G, Tan L (2014) Estimating the spatial distribution of soil moisture based on bayesian maximum entropy method with auxiliary data from remote sensing. Int J Appl Earth Obs Geoinf 32:54–66CrossRefGoogle Scholar
  21. Gesink Law DC, Bernstein KT, Serre ML, Schumacher CM, Leone PA, Zenilman JM, Miller WC, Rompalo AM (2006) Modeling a syphilis outbreak through space and time using the Bayesian maximum entropy approach. Ann Epidemiol 16(11):797–804CrossRefGoogle Scholar
  22. Hristopulos DT, Christakos G (2001) Practical calculation of non-gaussian multivariate moments in spatiotemporal Bayesian maximum entropy analysis. Math Geol 33(5):543–568CrossRefGoogle Scholar
  23. Hu J, Zhou J, Zhou G, Luo Y, Xu X, Li P, Liang J (2016) Improving estimations of spatial distribution of soil respiration using the bayesian maximum entropy algorithm and soil temperature as auxiliary data. PLoS One 11(1):e0146589CrossRefGoogle Scholar
  24. Jaynes ET (1957) Information theory and statistical mechanics. Phys Rev 106(4):620–630CrossRefGoogle Scholar
  25. Karlin S, Taylor HE (1975) A first course in stochastic processes, 2nd edn. Academic Press, New York, pp 495–502Google Scholar
  26. Koçak K (2002) A method for determination of wind speed persistence and its application. Energy 27(10):967–973CrossRefGoogle Scholar
  27. Lee SJ, Balling R, Gober P (2008) Bayesian maximum entropy mapping and the soft data problem in urban climate research. Ann Assoc Am Geogr 98(2):309–322CrossRefGoogle Scholar
  28. Martz HF, Waller RA (1994) Statistical methods for physcial science. In: Stanford JL, Vardeman SB (eds) Bayesian methods, vol 28. Academic Press, Inc., New York, pp 403–432Google Scholar
  29. Messier KP, Campbell T, Bradley PJ, Serre ML (2015) Estimation of groundwater Radon in North Carolina using land use regression and Bayesian maximum entropy. Environ Sci Technol 49(16):9817–9825CrossRefGoogle Scholar
  30. Modis K, Vatalis KI, Sachanidis C (2013) Spatiotemporal risk assessment of soil pollution in a lignite mining region using a Bayesian maximum entropy (BME) approach. Int J Coal Geol 112:173–179CrossRefGoogle Scholar
  31. Olea RA (1997) Understanding geostatistics. Course notes. Civil Engineering Department, University of Kansas, LawrenceGoogle Scholar
  32. Painter S (1998) Numerical method for conditional simulation of Levy random fields. Math Geol 30(2):163–179CrossRefGoogle Scholar
  33. Savelieva E, Demyanov V, Kanevski M, Serre M, Christakos G (2005) BME-based uncertainty assessment of the Chernobyl fallout. Geoderma 128(3):312–324CrossRefGoogle Scholar
  34. Şen Z (2001) Areal assessment of wind speed and topography with applications in Turkey. Renew Energy 24(1):113–129CrossRefGoogle Scholar
  35. Serre ML, Christakos G (1999) Modern geostatistics: computational BME analysis in the light of uncertain physical knowledge-the Equus Beds study. Stoch Environ Res Risk Assess 13(1–2):1–26CrossRefGoogle Scholar
  36. Serre ML, Kolovos A, Christakos G, Modis K (2003) An application of the holistochastic human exposure methodology to naturally occurring arsenic in Bangladesh drinking water. Risk Anal 23(3):525–528CrossRefGoogle Scholar
  37. Shannon CE (1948) A mathematical theory of communication. Bell Syst Tech J 27:379–423CrossRefGoogle Scholar
  38. Shi T, Yang X, Christakos G, Wang J, Li L (2015) Spatiotemporal interpolation of rainfall by combining BME theory and satellite rainfall estimates. Atmosphere 6(9):1307–1326CrossRefGoogle Scholar
  39. Süzek F (2007) Determination of wind energy potential of Turkey. M.Sc. thesis, İstanbul Technical University, TurkeyGoogle Scholar
  40. Xu Y, Serre ML, Reyes J, Vizuete W (2016) Bayesian maximum entropy integration of ozone observations and model predictions: a national application. Environ Sci Technol 50(3):4393–4400CrossRefGoogle Scholar
  41. Zhang C, Yang Y (2013) Spatial prediction of soil organic matter using Bayesian maximum entropy with histogram soft data. In: Bian F et al (eds) Geo-informatics in resource management and sustainable ecosystem. Springer, Berlin, Heidelberg, pp 674–683CrossRefGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2019

Authors and Affiliations

  1. 1.Department of Civil Engineering, School of Engineering and Natural SciencesAltınbaş UniversityIstanbulTurkey
  2. 2.Department of Meteorological Engineering, Faculty of Aeronautics and Astronauticsİstanbul Technical UniversityIstanbulTurkey

Personalised recommendations