REM Data Repositories
| Datasets/Assumptions | Description | Relevance for the methodology |
|---|---|---|
| Topographic data | The topographic data provided by Landfire is required to derive project-relevant topographic data. A maximum pixel size of 30 m x 30 m is required. | Topographic data is a relevant input for fire behavior modeling (see Section 6.3 of the methodology). Growth and yield modeling can be done aspatially and does not require topographic input. |
| Tree inventory | The latest iteration of the TreeMap dataset provided by USFS must be used as the underlying tree list. | TreeMap provides a tree-level model of conterminous U.S. forests at a 30m resolution. TreeMap data has to be updated by the project proponent to reflect the project’s starting year. Any disturbances that occurred on the project area between the TreeMap vintage year and the project start date must be fully captured (harvests, wildfire, fuel treatments, etc.) with appropriate, documented modifications to the TreeMap data. |
| Surface fuel models | Updated pre-treatment surface fuel models based on LANDFIRE 2016 have to be obtained and used from the First First Street Foundation. The process to obtain these data is as follows: 1) Send an email to the First Street Foundation Research Lab at [email protected] with REM Surface Fuels in your subject line and cc’ing [email protected]. 2) Include a shapefile for the area of interest (i.e., the project area) and a binding statement that these data will only be used for calculating wildfire emissions reductions for your project. Data requests come in on a rolling basis. A small team at First Street fields these requests, so please have patience with request turnaround times because there is often a data queue.
Post fuel-treatment fuel models and loads must be evidence-based, backed up by fuel models described by Scott and Burgan (2015) |
Surface fuel models and loads drive wildfire behavior and emissions. A realistic choice of surface fuel models and loads pre- and post-fuel treatment for both treated and untreated stands is a key driver for REM results. |
| Weather | Weather data must be derived from RAWS. | Weather data is needed to identify the project area and simulate stand-level wildfire impacts and wildfire behavior/spread. The project proponent has to use at least a 10-year average besides the requirements specified in Section 6.3 of the methodology. |
| Ignition points | Ignition points must be derived from USFS Spatial wildfire occurrence data for the United States. | Historical ignition points are required to model wildfire behavior. See Section 6.3. |
| Background forest management assumptions | Forest management modeled under both the baseline and project scenario on areas not receiving fuel treatments have to represent realistic management scenarios for the entire project area. This includes management of forests not actively participating in fuel treatment implementation. Defensible assumptions need to be backed by local evidence and/or literature (e.g., Table 6-6 in Eve et al. (2014)). | The choice of realistic scenarios that are representative of background management conditions is crucial for defensible growth and yield modeling under both the baseline and project activity to simulate changes to forest conditions. |
| Regeneration assumptions | Adjustments and a pulse of regeneration must be applied at every growth and yield modeling time step, along with a small-tree growth rate multiplier. These assumptions must be evidence-based and backed up by local documentation and/or literature. For California, Collins et al. (2011) is recommended as a source. | Certain FVS variants lack a forest regeneration model, leaving the user to input this information. This shortcoming can distort forest stand conditions as they are projected into the future based on user inputs which may be inconsistent or subjective. Depending on the understory conditions, projected canopy base height can increase rapidly, thereby greatly reducing the potential for crown fire initiation (Moody et al. 2016). |
| Delayed regeneration data | Assumptions on risk for delayed reforestation as well as carbon pools and fluxes for affected areas have to be evidence based and backed up by local data and/or literature. For California, see Buchholz et al. (2019) as a data source. | Proponents must quantify the area and emissions associated with project land that is projected to be temporarily or permanently converted from forestland to grass or shrubland following high severity fire over the crediting period. |
| List of approved fire behavior models | GridFire FlamMap ElmFire FSim |
The wildfire behavior model is used to calculate wildfire spread and the probability of a stand to burn. |
| Fire probability map | Annual fire probability risk based on Kearns et al. 2022, with following modifications:
|
Average annual fire risk within the project area is applied to modeling outcomes to provide probabilistic emissions estimates for FMU quantification. |