UC Science team response to SPLATs vs. DFPZs questions

UC Science team response to SPLATs vs. DFPZs questions by Kim_Ingram, at 3:33 p.m. on 17 March 2010,

George,

The following is the response from the science team to your questions and comments re. SPLATs vs DFPZs. If you have any follow up questions/comments, please feel free to use this discussion board so all interested parties may be informed.

3/17/2010

Reply to questions from George Terhune from the SNAMP FFEH Team.

1) The Meta-analysis was unpersuasive, due to the great differences among sites and treatments, which cannot be papered over with statistical manipulations.

Answer: The objective of the fire integration project is to compare the performance of disparate treatments. Thus the challenge is to construct meaningful measures of comparison. One of our attempts at comparison was with the normalized impact metric where we controlled for the percent of area treated. The resulting index does not have the intuitive meaning of the raw measures (e.g., changes in fireline intensity) but it does allow us to compare across sites. The statistical manipulations are not meant to paper over differences but to provide a valid means of comparisons.

2) Why has nobody yet done a direct comparison of the SPLAT strategy vs the DFPZ strategy, using the same terrain, fuel, weather scenarios, area of treatment, type of treatment, and modeling techniques, but varying only the pattern of treatments on the ground? After all, the fundamental difference between the SPLAT and DFPZ strategies is the pattern of treatment, and DFPZs are an exception to the Regional policy specifying SPLATs, and that exception was specifically made by Congress to "demonstrate the effectiveness" of the DFPZ strategy. It isn't enough to say "it’s effective." You have to say "How effective," "Compared to what?," and "How efficient, in terms of cost-effectiveness, effect on suppression cost and firefighter safety, and effectiveness at different levels of implementation, from a few percent of the landscape treated in early years, to perhaps 30 percent of the landscape treated after 10 or 15 years. What effect would each strategy have on suppression effectiveness and safety? If these questions aren't important enough to be investigated directly, what is it that would be more important?

Answer: The suggestion for an evaluation of multiple forest management options for one site is a good one. However it is not the goal of SNAMP. The role of the UC Science Team was defined by the federal and state agencies who signed the MOU that stated the goals of the project. The focus is on the forest management alternative approved in the 2004 Record of Decision.

The UC Science Team and the Pacific Southwest Research Station did make a commitment to collaborate in order to gain the most insight from the research efforts underway in the Sierra Nevada. The Fire Integration Study is part of this commitment. While the insights may frustrate in what they don’t do, they do support the general principle behind all coordinated landscape-level treatments. In short, their impact on modifying fire behavior across the landscape is more than the sum of the individual treatments.

A more fruitful exercise would be to compare the implemented DFPZ network in Meadow Valley to a theoretical SPLAT design for the same area. This comparison would not be perfect, but it may be worthwhile. The problem is who would design the SPLATs? A totally theoretical design (regular, louvered, rectangular treatments) would never actually happen on the ground (Collins et al. 2010). As was the case with the DFPZs in Meadow Valley the planned network differed considerably from the actually implemented network. One could imagine a similar situation with a planned SPLAT design and actual implementation. So, if a more realistic SPLAT design could be developed perhaps PSW could take on the question.

Collins, B.M., S.L. Stephens, J.J. Moghaddas, and J. Battles. 2010. Challenges and approaches in planning fuel treatments across fire-excluded forested landscapes. J. Forest 108(1):24-31.

This post is a part of the following discussions: