Blog‎ > ‎

Challenges to Forest Landscape Modeling Under Climate Change: The Road Forward: Avoiding Roadblocks

posted Aug 5, 2014, 9:17 AM by   [ updated Aug 5, 2014, 9:21 AM ]

Let’s assume that we all agree on the goals that I previously articulated.  How do we get there?  What are the potential roadblocks?  And how do we ensure that the journey was worth the effort?

First, to stretch the road metaphor to the breaking point, I assume that there is not a single road forward.  No one wants an all-encompassing forest model that purports to answer every question, to test every hypothesis, and at every scale.  Science simply doesn’t work that way.  Science is competitive and diverse and unbelievably dynamic.  Today’s little-known research project becomes tomorrow’s standard approach becomes the future’s outdated relic.  Failure is an option and is may be the best option for maintaining the requisite diversity of thinking and actors.  

Having said that, how do we guard against unnecessary and counter-productive model 'balkanization'?  How do we maximize model cross-fertilization?  An initial tactic will be to more frequently pursue opportunities for cross-scale model synthesis and comparison.  Model comparisons will identify areas of agreement and highlight where each model excels and which process representations should be promulgated.

In addition, we need to develop common libraries that will enable ready exchange of process representation and facilitate model synthesis.  A common spatial library, with a common suite of input and output interfaces (similar to the Geospatial Data Abstraction Library), an optimized grid array for managing data, and functions for neighborhood interactions, would be a good starting place; these are basic I/O functions that reflect a general data paradigm (raster) and do not reflect a particular scientific paradigm.  Clements and Gleason could both build a forest model on this foundation.  We have released the LANDIS Spatial Modeling Library as open source software for this very purpose.  

If models are built on a common platform, components could more readily be exchanged and ecologists would be free to focus more on the science and less on the underlying structure.  Although model platforms have been developed, e.g., SELES, they are often encapsulated within a higher-level programming language that limits process representation.  In contrast, program libraries written in a lower-level computer language (e.g., C or C++) can be operating system agnostic and easily wrapped for use with many other languages.  Success will require more widespread adoption of open-source principles (link to previous post).

Finally, forest simulations models are needed as tools in policy and decision support, particularly in relation to the effects of climate change.  Although the uncertainties remain numerous and large, management actions taken now will have consequences for the adaptation of many systems to climate change.  Forest models can play a critical role in evaluating adaptation responses.  Those of us that engage in forest modeling must therefore push harder to remove the barriers that often separate forest simulations from policy.  This will require improved communication about uncertainties, better access to output data, and ongoing and sincere engagement. As an example, see our projecting linking latest-generation forest climate science to management goals in Minnesota.  The recipe for being ignored is to continue doing what we have been doing without such improvements:  publishing scientific manuscripts and presenting to scientific audiences.  

I am fundamentally optimistic that we can achieve all of the above.  Many strides have been taken and progress continues on all fronts.  Even so, the financial resources required have risen while resources have shrunk. Creative solutions are required and ever-evolving.