Default Fuels Treatment / Disturbance Edit Rule Technical Documentation

This Documentation is presented in four parts:

Introduction
Methods
Results
Discussion

Introduction

In IFTDSS Users may choose to edit your landscape in two ways: By adding Default Fuel Treatment / Disturbance Edit Rules or by adding User Created Edit Rules. This enables users to update LANDFIRE fuel data layers with potential fuel treatments and test results with fire behavior and fire effects simulations. Users are required to use LANDIFRE “stock” disturbance type, severity, and time inputs in order to populate vegetation and fuel data for treated pixels.

LANDFIRE Look-up (LFLU) tables are used by IFTDSS when Users implement Default Fuel Treatment / Disturbance Edit Rules. LFLU tables are used to adjust the fuel attributes of cells, including: fuel model, canopy cover, stand height, canopy base height, and canopy bulk density, based on the type and intensity of treatment or disturbance and the time that has elapsed since that treatment or disturbance. This approach works by taking into account Existing Vegetation Type (EVT), Existing Vegetation Cover (EVC), and Existing Vegetation Height (EVH) and using the LFLU table as a crosswalk to account for changes in vegetation attributes, based on treatment or disturbance type, intensity and timing. That resulting vegetation is then crosswalked to one of the 40 fire behavior fuel models. More specifically, current vegetation conditions are extracted from each treated pixel and linked to a vector of post treatment existing vegetation and fuel conditions including FBFM40, CBD, CBH, CC, and CH and then mapped into each treated pixel and then used to “update” the fuel maps and to initialize fire behavior and fire effects simulations. In the continental US, only 36 of the 40 fuel models are used, the 4 fuel models that aren’t used are 109, 164, 201, and 202. As of now the LFLU rules are not available in AK or HI.

Fuel Model Look Up (FMLU) tables are used by IFTDSS when Users Implement User Created Edit Rules. These FMLU tables are generally based on the LFLU table transitions but do not utilize EVC, EVH, and EVT but rather transition fuel models to fuel models.

Methods

Methods Used to Develop Vegetation Transition Data

The LANDFIRE Vegetation Transition is a process that is used to couple disturbance and the vegetation upon which it has occurred, along with simulation models, in order to predict and map the resulting vegetation composition and structure. In addition, these data are leveraged in order to predict and map change in fuel attributes of the vegetation and the resulting fire regime attributes of the vegetation. Information from a variety of sources was used to inform vegetation transition assignments.

In forested EVTs, most of the transitions were developed using the Forest Vegetation Simulator (FVS) model, along with the Fire and Fuels Extension (FFE) model for fire simulations. These simulations were run on approximately 60,000 forested plots for all FVS variants nationwide. The plot data, contained within the Forest Vegetation Simulator (FVS) Ready Database (FVSRDB ) contain all attributes necessary for FVS simulations. These data also include predefined input tables used for initializing stand/plot information (StandInit and TreeInit tables) for FVS. More detailed descriptions and input data requirements can be found in the FVS User Guide (Rebain and others 2010). Following simulations, a database was derived from post-disturbance FVS/FFE outputs covering all FVS Variants at multiple severities and time-steps in the table below:

Disturbance type, severity, and time since disturbance.
Disturbance Type	Severity	Time Since Disturbance (TSD)	Time Since Disturbance (TSD)	Time Since Disturbance (TSD)
Fire	High	1 year	2-5 years	6-10 years
Fire	Moderate	1 year	2-5 years	6-10 years
Fire	Low	1 year	2-5 years	6-10 years
Mechanical Add	High	1 year	2-5 years	6-10 years
Mechanical Add	Moderate	1 year	2-5 years	6-10 years
Mechanical Add	Low	1 year	2-5 years	6-10 years
Mechanical Remove	High	1 year	2-5 years	6-10 years
Mechanical Remove	Moderate	1 year	2-5 years	6-10 years
Mechanical Remove	Low	1 year	2-5 years	6-10 years

These disturbances include fire, mechanical add, and mechanical remove. Each disturbance was simulated with three severities and results output at three different time steps. Severities include:

High severity where >75% above ground vegetation mortality is achieved in the simulation process
Medium severity where 25-75% above ground vegetation mortality is achieved in the simulation
Low severity where 0-25% above ground vegetation mortality is achieved in the simulation

Output time steps determine the Time Since Disturbance (TSD) and occur Initially after disturbance, 2-5 years post disturbance, and 6-10 years post disturbance.

Outputs from low and moderate severity simulations were synthesized in order to make them available for use in transition queries. The resulting disturbance effects on tree EVC and EVH were tabulated across all EVT’s and summarized using a majority solution derived for low, moderate, and high levels of aggregation of the EVT legend. This resulted in a single EVT/EVH outcome for every simulation type across all input EVC/EVH combinations. It was assumed that low and moderate severity disturbance did not change EVT.

A number of transitions were derived without the use of FVS or FFE. These transitions were instead based on a more heuristic approach. Stand-replacing events such as high severity fire and mechanical remove activities in forested EVTs were transitioned to an herbaceous or shrubland EVT’s with a cover and height appropriate for an early seral expression of that EVT and for that geographic location. In shrub EVTs , all fire severities were considered stand-replacing, so all burned non-forested polygons were replaced by an herbaceous EVT that would be found in that area. Mechanical treatments were treated similarly to fire disturbances and transitioned to an herbaceous community. Introduced annual grasses replaced some shrub-dominated EVTs in lowland areas (for example, Western US Great Basin and Columbia Plateau shrubland EVTs). In herbaceous EVTs, disturbed areas were not transitioned to different EVTs due to the fact that these communities rapidly reestablish themselves after disturbance. However, EVC and EVH were reset to values generally found after such disturbances.

LANDFIRE vegetation and fuel transition databases based on LANDFIRE Map Zone, EVT, EVC, and EVH provided the initial data for IFTDSS fuel transition data. The pre-treatment LANDFIRE vegetation data, including Map Zone, EVT, EVC, and EVH were required to link transitions to outcomes of fuel treatments. These data were developed in the fall of 2015 to provide the initial IFTDSS Fuel Treatment Updating version 1.0.

Methods Used to Develop LANDFIRE Fuel Transition Data

Similar to the vegetation transition process, the fuel transition process brings the fuel data to a more current time period with subsequent biennial updates and is based on the modeled change in vegetation (type, cover, height) through disturbance and natural processes. The same set of disturbances, severities, and TSD’s are used for fuel as are used for vegetation.

For surface fuel, transitions are based on the intensity of the disturbance and the time since the event. The surface fuel model was changed to reflect the change in surface vegetation from modeling and expert opinion inputs. Determination of the change in FBFM13/40 due to disturbance or natural processes was not based on information attained from Forest Vegetation Simulator. Instead, we looked at the relative change in fuel loading from the original LANDFIRE calibrated fuel model to help inform us with the surface fuel model transition due to disturbance. In some areas, the second time step since disturbance returned the fuel model to its original state due to the quick return of the vegetation. Vegetation in areas that were not included in the disturbance layer was modeled through growth simulation models.

An important distinction between vegetation, canopy fuel and surface fuel post disturbance assignments is worth highlighting. In determining post disturbance surface fuel model assignments there is no consideration of a fuel model call directly after a disturbance event (TSD 1). Since LANDFIRE is on 2 year update cycle, disturbance products are typically 2 years behind putting all new disturbances in TSD2. The assignments for surface fuels are expert opinion driven, based on the second growing season after a disturbance event. Essentially making the TSD 1 & 2 assignment the same due to LANDFIRE data production being 2 years behind the actual date. In some areas, the second time step since disturbance returned the fuel model to its original state due to the quick return of the vegetation. Vegetation in areas that were not included in the disturbance layer was modeled through growth simulation models.

Transitions in canopy fuel due to disturbance and succession were modeled using the FVS and FFE. The resulting disturbance effects on CC, CH, CBD, and CBH were tabulated across all EVT’s and summarized using a mean solution derived for low, moderate, and high levels of aggregation of the EVT legend. This resulted in a single canopy fuel outcome for every simulation type across all input EVC/EVH combinations.

Methods Used to Develop IFTDSS Fuel Treatment Updating version 1.0

Subsequent to this effort, the IFTDSS user process evolved greatly. New design features were implemented that allowed users to significantly edit the pre-treatment LANDFIRE fuel data once they had loaded their area of analysis. This included primarily editing the fire behavior fuel model (FBFM40) data, along with tree canopy cover (CC) and tree canopy height (CH) in order to improve simulation accuracies. The post-treatment fuel data, using the IFTDSS Fuel Treatment Updating version 1.0 data, now referred to as the LANDFIRE Lookup data or LF/LU, were then used to populate treated pixels. However, this process risks having a mismatch between the edited pre-treatment fuel data and the post-treatment fuel data. It was decided at this point to develop transition data based solely on pre-treatment FBFM40, CC classes, and CH classes and required the development of another version of the fuel transition data to be used in IFTDSS called IFTDSS Fuel Treatment Updating version 2.0.

Methods Used to Develop IFTDSS Fuel Treatment Updating version 2.0

A number of processing steps were necessary to mine V 1.0 databases and develop V 2.0 databases along with the associated completion dates. Basically, these steps started with unique combinations of Map Zone, pre-disturbance/treatment fuel model, stand cover class, and stand height class and selected the most commonly occurring post-disturbance/treatment fuel model and the average or mean stand cover, stand height, canopy base height and canopy bulk density values. This database was developed through a hierarchical process in order to leverage disturbance and fuel treatment outcomes nationally, regionally, and sub-regionally so all possible outcomes are addressed and data populated all possible combinations of the pre-treatment fuel data.

Before making a final data deliverable package process checks were implemented to ensure the canopy assignments did not create erroneous values. A general check was first implemented to ensure canopy structure and canopy fuel values aligned in a logical manner. If the canopy cover was less than 10% or canopy bulk density was equal to zero it was assumed there was not enough canopy present to carry canopy fire and subsequent canopy characteristics (CC, CH, CBD and CBH) were masked out. Finally, to ensure no illogical canopy base height values were assigned a check was implemented to check for and reduce values to 2/3 the canopy height if greater.

Disturbance Details Used to Inform LFLU and FMLU for use in IFTDSS

Below are the descriptions of each of the Default Fuel Treatment / Disturbance Edit Rules and the assumptions made when modeling the reaction of stands in FVS to inform the LFLU tables used for adjusting the fuel attributes. The FMLU tables are generally based on the LFLU table transitions but do not utilize EVC, EVH, nor EVT. Rather, they transition fuel models to fuel models.

Thin: Slash Removed

Light Thinning; Pile Burning: The first phase of this treatment assumes an understory thinning treatment, applied to <25% of the area; thinning the stand to 80% of current density by thinning up to a 6" diameter. Subsequent pile burning assumes thinned material is removed by burning 70% of the area at 80% consumption for each pile. An additional 20% mortality is assumed in <5” diameter size classes as a result of thinning and subsequent pile burning.
Moderate Thinning; Pile Burning: The first phase of this treatment assumes an understory thinning treatment, applied to 25-75% of the area; thinning the stand to 35% of current density with no upper diameter limit. Subsequent pile burning assumes thinned material is removed by burning 70% of the area at 80% consumption for each pile. An additional 20% mortality is assumed in <5” diameter size classes as a result of thinning and subsequent pile burning.

Thin: Slash Remains

Moderate Thin; Lop and Scatter: This treatment assumes <25% of the area affected by an understory thinning treatment cutting 55% of all material up to a 6” diameter and leaves it on site (lop and scatter).
Moderate Thin; Masticate: This treatment assumes 25%-75% of the area affected by an understory thinning treatment cutting 75% of all material up to a 6” diameter and leaves it on site as masticated material.
Heavy Thin; Masticate: This treatment assumes >75% of the area affected by an understory thinning treatment that cuts 90% of all material up to an 8” diameter and leaves it on site in the form of masticated material.

Clear Cut

Clearcut and Broadcast Burn: This treatment consists of removal of 100% of the overstory followed by broadcast burning across >75% of the area.

Wildland Fire

Low severity fire: Low severity wildfire or prescribed burn in natural fuels consisting of a patchy burn of <25% mortality of the above ground vegetation over 45% of the area.
Moderate severity fire: Moderate severity wildfire or prescribed burn in natural fuels; mixed severity fire with 25-75% mortality of the above ground vegetation over 75% of the area.
High severity fire: High severity wildfire or prescribed burn in natural fuels mixed severity fire with stand replacement fire in some areas; mortality of above ground vegetation from 75-100% over 95% of the area.

Time Since Disturbance

1 Year - This assumes that the fuel treatment or disturbance occurred the previous year. This time period assumes a minimum reaccumulation of fuel post disturbance or treatment but recognizing that this reaction will vary depending on EVT, disturbance or treatment intensity, and local factors of the LANDFIRE zone.
2-5 Years - This time period assumes a median of 4 years. Utilizing FVS the LANDFIRE team utilizes the median of this period to model vegetation recovery and fuel accumulation for the purposes of informing the local workshops of local experts to determine an appropriate new EVT and resulting fuel model.
6-10 Years - This time period assumes a median of 8 years. Utilizing FVS the LANDFIRE team utilizes the median of this period to model vegetation recovery and fuel accumulation for the purposes of informing the local workshops of local experts to determine an appropriate new EVT and resulting fuel model.

Results

For each disturbance type, severity, and TSD, an Excel table was constructed through a series of database queries. The table below shows the structure of those tables.

Field names and descriptions.
Field Name	Field Description
MZ	MapZone
VDIST	Concatenated disturbance type/severity/TSD code
EVT01	Pre-disturbance EVT
EVH01	Pre-disturbance EVH
EVC01	Pre-disturbance EVC
EVT12	Post-disturbance EVT
EVT_Fuel	Post-disturbance EVT recoded to EVT fuel
EVH12	Post-disturbance EVH
EVC12	Post-disturbance EVC
FBFM13	Fire Behavior Fuel Model 13
FBFM40	Fire Behavior Fuel Model 40
CanCov	Canopy Cover
Can Hgt	Canopy Height
CBD	Crown Bulk Density
CBH	Crown Base Height

All in all, 27 individual tables were derived for this effort using these methods. A number of caveats exist that one should be aware of while using these data. Using each one of these 27 tables created above, another set of Excel tables was constructed through a series of database queries. The table below shows the structure of those tables:

Field names and descriptions.
Field Name	Field Description
MZ	MapZone
VDIST	Concatenated disturbance type/severity/TSD code
FBFM13	Pre-disturbance Fire Behavior Fuel Model 13
FBFM40	Pre-disturbance Fire Behavior Fuel Model 40
CanCovCls	Pre-disturbance Canopy Cover Class
CanHgtCls	Pre-disturbance Canopy Height Class
FBFM13	Post-disturbance Fire Behavior Fuel Model 13
FBFM40	Post-disturbance Fire Behavior Fuel Model 40
CanCov	Post-disturbance Canopy Cover
CanHgt	Post-disturbance Canopy Height
CBD	Post-disturbance Crown Bulk Density
CBH	Post-disturbance Crown Base Height

Discussion

It should be noted that although developed from LANDFIRE base data, logic, and databases, it is likely that post disturbance vegetation and fire behavior characteristics of the IFTDSS/LFLU will differ from those of standard LANDFIRE data. These differences will occur due to process and logic framework that is dissimilar between the two systems. There are at least three fundamental differences between LANDFIRE and IFTDSS fuel production and resultant fire behavior characteristics.

In the process of building the IFTDSS assignment databases, cover and height bins were abbreviated into 3 groups for cover and 3 groups for height, condensing current LANDFIRE. In current LANDFIRE production canopy cover (CC) is derived by taking the midpoint of EVC bins, which are 10% increments (0 to 100%). Canopy height (CH) is derived by taking midpoint of EVH bins (5 bins for treed pixels). The canopy attributes of LANDFIRE CC and CH play a pivotal role in the calculations of canopy base height (CBH) and canopy bulk density (CBD). The differentiation from the grouping of these classes creates the likelihood combinations of CC, CH, CBH, and CBD will not match those of current LANDFIRE. For instance, it is possible in the IFTDSS tables for an area to undergo a treatment and end up with a higher CC than the original LANDFIRE data due to this grouping. There will also be other fire behavior differences caused by the grouping of canopy values:

Fire behavior characteristics will differ considerably due to increases and decreases in CC & CH and the resultant variation in wind reduction factors (especially in the moderate cover ranges) which effects the amount of wind interacting with the surface fire.
The ability for crowning to initiate will be considerably different between the LFLU table canopy values and standard LANDFIRE. In the LFLU CBH and CBD are derived directly from FVS outputs whereas in LANDFIRE they are calculated from EVT, EVC, and EVH on per pixel basis.

The IFTDSS/LFLU process differs from the original LANDFIRE data in terms of the potential for crown fire activity in certain vegetation types. In the LANDFIRE fuel and fire behavior process the tree species that are reluctant or resistant to crown fire initiation or propensity (often time’s deciduous tree species) can be addressed through a canopy guide which makes them unable to crown. There is no such system built into the IFTDSS/LFLU.
Although EVT, EVC, EVH, Bps, and Disturbance are inherent in the development of the original LANDFIRE fuel rulesets to achieve a surface fire behavior fuel model assignment, that is not necessarily the case in the LFLU tables. The processes are fundamentally different in that the pathway to the post disturbance fuel model are different. Within original LANDFIRE data the vegetation type, the structure characteristics, and the non-disturbed (beginning) fuel model are the basis for determining the change to the post disturbance fuel model assignment. In the IFTDSS/LFLU just the beginning fuel model and disturbance is the pathway to the post disturbance fuel model assignment (Table 2).
Because the LANDFIRE surface fuel models reflect a site condition 2 growing seasons post disturbance (TSD 1 and 2), the resultant fuel model assignment is always a burnable fuel model. IFTDSS users may want to reflect a site condition within the first 2 years post disturbance (TSD 1) that may reflect a non-burn surface fuel model.