Blog‎ > ‎

Challenges to Forest Landscape Modeling Under Climate Change (Part 1)

posted Aug 18, 2012, 8:07 AM by Robert Scheller   [ updated Aug 18, 2012, 8:12 AM ]
Forest landscape modeling has seen dramatic improvements since the first models were conceived nearly 30 years ago.  Among others, the spatial resolution of the models has increased considerably as has the number and complexity of the ecological processes that are incorporated into the models.  For example, 10 years ago only one or two processes - typically either fire, wind, or harvesting - were included in any given study.  Today we can access large libraries of many processes from which to choose.

Perhaps rightly so, then, many forest ecologists perceive forest landscape modeling as a success story.  These models are providing important information to land managers about adaptive strategies to cope with the effects of an altered or changing environment on forest resources.

In spite of our successes major challenges remain for scientists trying to represent the dynamics of very complex entities within a computer model.  Some challenges are not unique to climate change but are inherent to any attempt to capture the dynamics of highly complex systems.  For example, many (or most?) ecological systems are constantly in flux (‘non-stationary’) due to human pressures and such systems are notoriously difficult to model.  As a result of this constant flux, the common assumption of quasi-equilibrium is not appropriate given environmental changes that operate faster than typical landscape-ecological dynamics. Other universal challenges include non-linear dynamics, higher-order interactions among processes, and validation of the myriad components.

Other challenges are unique to climate change :  
  • The large inherent uncertainty associated with climate change (we cannot anticipate everything that could happen in a climate-altered future).
  • Our lack of consistent data about the future (for example, there are dozens of projections of climate change).
  • Parsimonious representation  within our models - being mechanistic where we need to be and not where we don’t.
  • Validating projections of climate change - how do we know if a model is working when we’re modeling the distant future?
  • Model formulation: our models may be poorly designed or calibrated for ecological processes that are climate or weather driven.  For example, many models are inappropriate because they were not originally designed to represent the complex dynamics associated with climate change.
The biggest challenge may be perceptual.  Is ecological forecasting a reasonable goal considering the large uncertainties combined with high rate of change?  If not, what are our goals?  Can we learn about the future fast enough to positively influence policy choices?

Further thoughts on these challenges in the coming months.

- Robert Scheller