METLA - Heikkinen, Juha

Selected publications

Original research papers

[O1] Heikkinen, J. and Högmander, H. (1994): Fully Bayesian approach to image restoration with an application in biogeography. Applied Statistics 43, 569-582. [  abstract  ]

• Bayesian image analysis techniques applied in the estimation of biogeographical ranges from incomplete presence data in quadrants, taking into account spatially heterogeneous research intensity

[O2] Heikkinen, J. and Penttinen, A. (1999): Bayesian smoothing in the estimation of the pair potential function of Gibbs point processes. Bernoulli 5, 1119-1136. [  abstract  ]

• an almost nonparametric method for estimating the pairwise interactions in regular spatial point patterns

[O3] Arjas, E. and Heikkinen, J. (1997): An algorithm for nonparametric Bayesian estimation of a Poisson intensity. Computational Statistics 12, 385-402. [  abstract  ]

• Poisson process on the real line (time); the technique is generally applicable in curve estimation (see e.g. [R4])

[O4] Mäkisara, K., Heikkinen, J., Henttonen, H., Tuomainen, T., and Tomppo, E. (1997): Experiments with imaging spectrometer data in large-area forest inventory context. In Proceedings the Third International Airborne Remote Sensing Conference and Exhibition. Development, Integration, Applications & Operations. 7-10 July 1997, Copenhagen, Denmark. Volume II. pp 420-427. [  abstract  ]

• attempts to utilise high resolution (1.6 m pixels); including experiments with spatial (textural) features

[O5] Heikkinen, J. and Arjas, E. (1998): Non-parametric Bayesian estimation of a spatial Poisson intensity. Scandinavian Journal of Statistics 25, 435-450. [  abstract  |  simulated data sets  |  related c code  |   S-Plus functions  |   documentation for the latter  ]

• spatial version of [O3]; generally applicable in surface estimation (see e.g. [R4])

[O6] Heikkinen, J. and Arjas, E. (1999): Modeling a Poisson forest in variable elevations: a nonparametric Bayesian approach. Biometrics 55 738-745. [  abstract  |  c code and S-Plus functions  ]

• utilising methods developed in [O3] and [O5] for estimating the response of plant abundance to changes in environmental variables; involves simultaneous nonparametric estimation of the response curve and baseline intensity surface and interpolation of the surface of the explanatory variable values
• Revised version of Heikkinen and Arjas (1997): Nonparametric Bayesian estimation of a spatial Poisson intensity using a partially observed concomitant variable.

[O7] Katila, M., Heikkinen, J., and Tomppo, E. (2000): Calibration of small-area estimates for map errors in multisource forest inventory. Canadian Journal of Forest Research 30 1329-1339. [  abstract  ]

• a simple statistical correction to the multisource estimators of land-use class areas, to compensate for the map errors, and
• how to extend the correction to other multisource estimators, e.g. that of the growing stock volume.

[O8] Kühlmann, S., Heikkinen, J., Särkkä, A., and Hjorth, U. (2001): Relating abundance of ground vegetation species and tree patterns using ecological field theory. In K. Rennolls (editor): Proceedings of IUFRO 4.11 Conference: Forest Biometry, Modelling and Information Science. [  abstract  ]

• earlier studies of single-tree influences to vegetation have been in relatively small scale (homogeneous study areas)
• here, we wanted to see if useful results could be obtained from a nationwide inventory with small sample plots
• the major issues to deal with were the small size of individual sample plots and the stong effect of variable stand-level conditions

[O9] Heikkinen, J. and Henttonen, H. (2002): Re-measured relascope plots in the assessment of inventory update methods. In J. Heikkinen, K. T. Korhonen, M. Siitonen, M. Strandström and E. Tomppo (editors): Nordic Trends in Forest Inventory, Management Planning and Modelling. Proceedings of SNS Meeting in Solvalla, Finland, April 17-19, 2001 113-118. The Finnish Forest Research Institute, Research Papers 860.

• how to account for the extra random component, resulting from relascope sampling, in the plot-level difference between predicted and observed time 2 volumes
• more robust alternative to the method of Roesch et al. (1993, Forest Science 39, 66-77).

[O10] Mäkipää, R. and Heikkinen, J. (2003): Large-scale changes in abundance of terricolous bryophytes and macrolichens in Finland. Journal of Vegetation Science 14 497-508. [  abstract  ]

• analysing statistical significance of changes from forest inventory data taking into account spatial correlation, systematic sampling, partial re-measurement and multiple testing

[O11] Lehtonen, A., Mäkipää, R., Heikkinen, J., Sievänen, R. and Liski, J. (2004): Biomass expansion factors (BEFs) for Scots pine, Norway spruce and birch according to stand age for boreal forests. Forest Ecology and Management 188 211-224.

• statistical contribution is related to the joint analysis of model and sampling errors in inventory data

[O12] Taskinen, I. and Heikkinen, J. (2004): A nonparametric Bayesian method for assessing uncertainty in thematic maps of forest variables.

• an alternative to the k-nn method, allowing for model-based assessment of uncertainty
• based on partition modelling (Denison et al.: Bayesian Methods for Nonlinear Classification and Regression, Wiley 2002), an approach similar in spirit to that of ([O3], [O5], [O6], [R3], [R4], [R7]) but computationally much less expensive.

[O13] Kühlmann-Berenzon, S., Heikkinen, J. and Särkkä, A. (2005):