One of the world’s oldest and largest timberland and agriculture investment management organisations had a particular problem that required our expertise.
The company manages more than 1.7million acres of forestry on four continents, representing assets and commitments of $2.5 billion, including investments from several pension funds. Decisions affecting its forests can have an important impact on how much pensioners might receive in the future.
Part of an area of forestry under its management was not growing as well as those surrounding it and the company wanted to know if a climate factor was inhibiting growth patterns.
The project differed from other forest inventory modelling cases, as the company provided us with data for 10 forestry and climate ground plots. It wanted us to provide historic rainfall, temperature and humidity statistics over the previous three years to generate an ecosystem model against which we could to calculate growth rates over an area of 427ha.
The project was made more complicated by the fact that the amount of climate data available to us was partial and covered a small geographical area and we required to use machine learning to predict accurate temperature, rainfall and humidity data for the relevant area.
Within a short timescale we were able to produce a forest inventory that factored in climate variables to demonstrate with a high degree of certainty that the lack of growth in the affected area was not caused by climate but rather by other factors.