(Chris J. Cieszewski)
The Fiber Supply Assessment position at the Warnell School of Forest Resources is budgeted for 100% research. Fiber Supply Assessment consists of estimating current and future states of forest inventory. Research on estimation of current forest inventory is mainly related to sampling and study of statistical estimators, which for the state’s inventory are exclusively handled by the USDA Forest Service. Research on estimation of future forest inventories is mainly related to studying methods relating to: estimating site productivity; and projecting with special inventory projection models various stand attributes into the future, which enables the estimation of the future inventories’ characteristics. The future inventories are computed from: i) the site productivity estimation; ii) the estimation of the future inventories’ characteristics; and iii) various assumptions about future removals.
Focusing on research of statistical estimators has not been pragmatic so far, because the institution that has monopoly on the state’s inventory measurements and sampling design expressed no interest in any of its results, and both the Dean and the USDA Forest Service have discouraged such efforts. Therefore, the Fiber Supply Assessment program has been mainly focusing on development of new improved methods for site productivity estimation and for developments of improved models for inventory projections. These include mathematical methods for derivation of initial condition difference equations called also dynamic equations, and statistical methods for stochastic estimation of their parameters.
In these areas I have had outstanding accomplishments as measured by solving vital problems, impact on operational implementation by the forest industry members, and by amount and quality of publications. I have developed breakthrough concepts in the area of implicitly defined initial-condition difference equations and published articles on them in journals such as Forest Science and The Canadian Journal of Forestry Research, which are the two most prestigious journals in the research of forestry and forest biometrics. The most significant contribution to forest biometrics is the development of the Generalized Algebraic Difference Approach (GADA) (Cieszewski and Bailey 2000, see CV-3a.v.). The Generalized Algebraic Difference Approach to theory-based derivation of self-referencing (inventory-defined) solves major problems that had been limiting the effectiveness of forest growth and yield modelers for nearly 70 years (at least since Schumacher 1935). I also specialize in parameter estimation (mixed effects modeling), which is at the cutting edge of statistical theory and addresses problems that have challenged foresters for more than 25 years (problems discussed already in 1974 by Bailey and Clutter).
The published articles and the current work are fundamental to site productivity estimation and inventory projections. Counting publications in press, I have published 23 referred articles, 25 non-refereed reports, and 4 abstracts (see CV-3a) related to the Fiber Supply Assessment concepts. Among the refereed publications there were 18 first-authored manuscripts: 7 in the leading international journals dedicated to forestry and forest biometrics, 4 in international journals, and 7 in other refereed outlets. The research on site productivity estimation and inventory projection functions was also numerously presented at regional, national, and international conferences (see CV-3h and CV-4D).
The published work has had a substantial impact on various levels of influence. At the Warnell School of Forest Resources faculty members use these methods in their research and in their courses, graduate students use them methods in their research and the PMRC uses these methods in place of the more traditional methods that were in use prior to my appointment at UGA.
Outside of the school, Weyerhaeuser Inc. has implemented the new methods in their operations, and committed $10,000 as seed money for a consortium supporting this research. My work has become required reading in graduate studies at Oregon State University forestry school. It has been used by Wayerhauser biometricians in Olimpia, WA, to make major advances in modeling of Douglas fir stands.
My research on new methods for stand inventory projections and fiber supply assessment is also recognized and applied in Canada, Sweden, Estonia, Portugal, and Poland and has been much cited nationally and internationally (see CV-3d). My models have been found to be the most suitable modeling tools in Estonia and Sweden as established by (see CV-3d) Elfving and Kiviste (1997 For. Ecol. Manag. 98: 125-134), Eriksson et al. (1997 Scand J. For. Res. 12: 149-156), Kiviste (1998 Tallinn. 191-196), and Johansson (1999 Scand. J. Forest. Res. 14: 441-453).
I have reviewed dozens of manuscripts for journals (see CV-4A), forest industry members (see CV-4B), and individual scientists (for example, prof. Carmean, Lakehead University, Thunder Bay, ON, Canada). I have served three times as Guest Associate Editor for Forest Science (see CV-3g), was appointed along with Prof. H. Burkhart, distinguished scholar, as the Guest Editor of the Journal of Mathematical Modelling and Scientific Computing (see CV-3a.v.2001); and was Editor of the proceedings of the 1st International Conference on Measurements and Quantitative Methods and Management (see CV-3a.v.2000).
I fulfilled my outreach obligations and contributed over and above the duties to the service mission of the UGA through: public speaking; web pages development for reporting research results and its progress; presentations to PMRC, CAPPS and other professional meetings, organizations; organization of chairing the First International Conference on Quantitative Methods and Management where all participants gained in total over 600 hours of continued education credits, and other outreach functions (See CV-4C and 5).
I have guided four 3-semester-hour graduate courses in directed studies (see CV-2) for Ph.D. and M.S. students in the areas of stochastic regressions, derivation of inventory projection models, and comparisons of three types of dendrometers used for diameter measurements for research and inventory purposes. I have also taught invited lectures and seminars at the University of Georgia (see CV-2), and at other universities and in other countries (see CV-4D) and contributes to teaching locally, nationally, and internationally with new methods that I have developed, or improved, in the fields of forest biometrics and forest inventory data analysis.
My 3rd year review recommended improvements in the amount of PI-grants, number of graduate students, and broadening the focus of my research to include elements of applied analysis of regional data. Since then I have fully satisfied all the recommendations and made major improvements in all the identified areas.
Accordingly, prior to the 3rd year review I had been PI on 2 grants for $3,000 and 3 un-funded proposals for $330,000, and Co-PI on 2 grants for $180,000. Since then I have been PI on 4 grants for $320,000 and 3 un-funded proposals for $387,000, and Co-PI on 2 grants for $162,000.
Furthermore, prior to the 3rd year review I had had no graduate students, served on 2 Ph.D. committees, and guided one 3 hours graduate course in directed studies. Since then I have had 6 graduate students, served on 9 Ph.D. and M.S. committees, and guided four 3-hour graduate courses in directed studies for total credit of 21 semester-student-hours.
Finally, with respect to the recommendation of extending my
research program to applied research I have expanded my study including applied
analysis of the state/regional/national inventory data and operational
assessment of current and future fiber supplies in Georgia. As a result, the current research of my
program includes: i) analysis of GIS and forest inventory data on the state and
the regional level; ii) research on new methodologies for derivation of
inventory projection models and analysis of stochastic parameter estimations
for these models.