# An Introductory Tutorial on Common Methods for Determining Fire Frequency

Joseph Fall

Introduction

The purpose of this tutorial is to present a clear and comprehensive description of common, published methods for computing historical fire frequency. This tutorial is structured as a set of interactive modules with worked examples programmed in Excel 97 for Windows. Related topics are connected with hyper-links, which are used to traverse the tutorial. The following topics are covered in the tutorial:

• An introduction to fire regimes, fire frequency, the evidence and data used to estimate fire frequency, measures of fire frequency, and fire frequency models, including a glossary and bibliography.
• A tutorial on the Natural Fire Rotation method for estimating fire frequency.
• A tutorial on the Poisson model of fire frequency.
• A tutorial on the Fire Return Interval method for estimating fire frequency.
• A tutorial on potential pitfalls of working with Fire Interval Data.
• A tutorial on the Negative Exponential and Weibull Fire Frequency Models.
• A tutorial on the Fire Cycle method for estimating fire frequency.
• A tutorial on the new Maximum Likelihood Estimator method for the estimating the Fire Cycle
• A suite of simulation tools for designing and running aspatial fire frequency simulations. (This module is also available separately.)

It is important to note that this tutorial is not meant to advocate any particular method, or to propose new or emerging approaches to fire frequency reconstruction. Nor is it sufficient to equip people with all the tools they would require to undertake a fire frequency analysis. It is simply a review of published material that is meant to help aid comprehension of that material.

The tutorial has a module that allows you to design and run aspatial simulations of a variety of fire frequency models. These simulation tools present two distinct approaches to modeling fire frequency. The first is a literal translation of the Weibull / Negative Exponential type of model, run as a simulation. The second defines fire frequency based on the area return interval and a distribution of fire sizes. In either case, you are provided with a flexible set of parameters that let you create a variety of scenarios, and directly explore the consequences of some of the assumptions made by these models. While the simulation runs, it displays graphs of the current landscape time-since-fire (or age-class) distribution and the current fire-interval distribution. This module is packaged with the tutorial, and is also available separately.

Motivation and Scope

I was originally motivated to produce this tutorial because I found it difficult to compare different methods for computing fire frequency presented in the literature. I also found that working with these published data sets in a consistent framework helped me to grasp some of the concepts and computations involved in estimating historical fire frequency. While the tutorial covers only a single aspect of fire history reconstruction (i.e., the mechanics of computing fire frequency), I hope that it will serve to make this portion of the science accessible to others who do not make it their career. In this section I will further outline the objectives, relevance, and scope of the tutorial.

I have two primary objectives for this tutorial:

1. to review the published fire history literature and determine the methods, underlying models, and calculations being used, along with the statistical and ecological assumptions and interpretations being made in fire frequency reconstructions
2. to synthesize and convey this material in a format that will help a novice reader understand what has been done, how it was done, and why it was done that way.

This tutorial was particularly relevant in B.C. at the time of its development because of new initiatives and projects underway at the Ministry of Forests. In particular, the Forest Practices Code Biodiversity Guidebook uses a rough fire cycle analysis employing the negative exponential fire frequency model to estimate the seral stage distributions required to meet biodiversity objectives. Any forest manager who wishes to understand how these distributions were derived, and to evaluate the applicability of this model, needs to have a clear understanding of the fire cycle method, the negative exponential model, and their assumptions and interpretation. A project currently underway in the Nelson Forest Region provides a good example . The authors of that study use a fire cycle analysis based on the Provincial forest cover maps (the FIP/SEG database) and, based on this analysis, suggest that the seral stage requirements for the region should be altered and the Annual Allowable Cut (AAC) for the region should be increased. It is very important that the people who will review this material and make decisions about these proposed changes should have access to a clear presentation of the underlying models that were used in the original analysis.

Finally, the scope of this tutorial is limited to reviewing published methods for computing fire frequency. An understanding of these computations is only one of the many tools and skills required to reconstruct a fire history. In this respect, it would be useful to include this fire frequency tutorial within a larger package of fire ecology and fire behaviour tutorials. However, for the time being it is important to keep in mind that the scope of this tutorial is limited, and that it is designed only to provide access to understanding fire frequency analyses. It is not meant to advocate any particular method, nor is it meant to propose new or emerging approaches to fire frequency reconstruction. The tutorial is not sufficient to equip people with all the tools they would require to undertake a fire frequency analysis.