MCM Forest Management | Multi-Objective Modeling of Carbon, Economy, and Water Conservation

This project addressed Problem E of the 2022 MCM/ICM. The task was to model carbon sequestration by forests and wood products, design a forest-management plan that balances harvesting with other ecological values, apply the model to a specific forest, and explain the recommendation to the local community. The team paper combined carbon sequestration, water conservation, and economic value in a CWE model, then applied it to Larix olgensis plantations and natural secondary Quercus mongolica forests in eastern Liaoning, China. I was primarily responsible for modeling and contributed to a limited portion of the paper writing; exact submodel and section boundaries were not further specified.

Forest management is not a single-objective problem. Allowing trees to grow supports carbon sequestration, while appropriate harvesting can store carbon in wood products and create economic value; harvesting also changes vegetation structure and water-conservation capacity. These objectives use different units, time horizons, and directions, so direct comparison can be misleading. Tree species, age classes, stand density, wood carbon content, cultivation costs, terrain, and climate introduce further complexity that must be controlled through explicit assumptions when data are limited.

The project had four goals: estimate carbon sequestration by forests and wood products over time; build a decision model balancing sequestration, harvesting economics, and water conservation; demonstrate how the model could be applied to a selected forest; and translate the conclusion into a nontechnical article for the local community. This page presents the modeling process, team results, and confirmed award supported by the available materials. It does not imply real-world deployment or unverified independent ownership.

The modeling inputs covered forest-stock growth, biomass across age classes, average organic carbon content, wood density and carbon content, and cultivation, management, and harvesting costs. The water-conservation analysis organized five factors: rainfall, soil texture, land cover, vegetation, and landform. The application focused on representative Larix olgensis plantations and natural secondary Quercus mongolica forests in eastern Liaoning. This section describes data and source investigation for modeling rather than user research; source datasets, code files, and the part of data preparation I personally owned are to be added.

The solution used a layered structure. The carbon model separated living trees from wood products: a Richards equation fitted forest-stock growth for living trees, and biomass, average organic carbon content, and age-class area were used to estimate sequestration potential; wood products were modeled from wood carbon content and density. The economic model combined harvesting revenue with accumulated cultivation, management, and labor costs. An AHP model weighted five water-conservation factors and checked judgment-matrix consistency. The normalized objectives were then combined into a weighted CWE decision function.

The paper established assumptions, notation, and submodels before calculating the AHP judgment matrix, factor weights, and consistency ratio. Carbon sequestration, economic value, and water-conservation objectives were normalized and combined with equal weights of 0.33. For the regional application, forest area, species structure, age classes, carbon content, and wood density from eastern Liaoning were substituted into the model, with harvesting quantity and interval as the main decision variables. I primarily handled modeling and contributed to a limited portion of the paper writing; the programming language, solver, visualization tools, code modules, and exact derivations I owned were not further specified.

For the eastern Liaoning case, the paper reported a management result of harvesting 450 trees per hectare every 30 years and used that schedule to discuss long-term sequestration by forests and wood products. The team also produced a 24-page English submission and a nontechnical community article titled “Is Logging Necessarily Bad?” explaining why appropriate harvesting can be part of forest management, and ultimately received an Honorable Mention. This was an output of the paper’s model, not evidence of real-world deployment or independent external validation; detailed reviewer feedback has not been added.

The CWE model placed environmental, economic, and socially relevant factors in one decision framework and made tradeoffs visible, but its output depends on simplifying assumptions and experience-based parameters. The body of the paper used equal objective weights, whose justification could be strengthened. The application focused on two tree types and did not fully represent climate change, biodiversity, or other human activities. Future work could add parameter sensitivity, alternative weighting scenarios, uncertainty ranges, cross-region validation, and reproducible data-processing and optimization code. Exact submodel ownership, paper-section boundaries, and a fuller post-competition retrospective can be added when more information is available.