We propose a rapid response project to collect fire behavior, fire effects, and fuels data from five 2003 active 2004 wildfires across the US. It is critical that field and remotely sensed data be collected soon (two weeks to the first growing season) after wildfires are burning if we are to understand the interactions and spatial variability in fire effects, fuels, fire behavior, local weather and topography and to assess the accuracy of current and alternative image analyses for remote sensing of burn severity. This most directly addresses Task 2.1 (AFP 2003-2, Task 1) by quantifying the relationships between pre-fire fuel structure, fire behavior and post-fire effects, and also Tasks 4.1 and 1.1. On all ten fires, we will collect field data on pre and post-burn fuels, fire behavior, fire effects on soils and vegetation composition and structure, and compare them to pre and post-burn Landsat imagery. If possible, we will select two of the five wildfires we sample each year from locations for which we have extensive preburn vegetation and fuels information. On those sites, we will ensure that we sample exactly the same points as where another rapid response team proposes to collect detailed instrumental data on fire behavior and heat flux, as well as some preburn data. For two of those sites we will also obtain hyperspectral and lidar imagery for in-depth analysis. Our rapid response team will work closely with and share data and results with Fire Use, Incident Command, and BAER teams. Our efforts will complement ongoing research and management applications by comparing alternative remote sensors and analysis approaches across a diversity of soils, vegetation, and fire conditions, and by explicitly linking fire behavior, fuels and fire effects to quantitative indicators of burn severity, that can be assessed in the field, predicted from fire effects models and mapped. We will compare spatial variability of preburn site and vegetation conditions, heat flux, fire effects and satellite imagery. Co-PI Robichaud and others are studying the spatial variability of fire effects and evaluating hyperspectral, AVIRIS and SPOT remote sensing tools for mapping fire effects on the Hayman and Williams fires. Co-I Jam and others are predicting post-burn fire effects (soils and tree effects) from pre-burn forest structure based on extensive sampling of areas burned in the northern Rockies in 2000 and 2001. Collaborator Key and others are validating delta Nonnalized Burn Ratio (dNBR) across a wide range of sites. We will also organize a research and applications workshop for researchers and applications specialists from multiple federal agencies to synthesize and recommend methods for quantitative field measurement and remote sensing of burn severity. We will work to develop a rapid yet consistent burn severity mapping approach that is applicable to different types of imagery (depending on which is available at the right time). We will share our results at BAER training, on the FIREMON (https://www.frames.gov/partner-sites/firemon/firemon-home/) and FRAMES (www.frames.gov) websites, and with the USFS Remote Sensing Applications Center staff working to improve their procedures for mapping fire effects. The spectral library from all sites will also be available on FRAMES. The data will be useful to those developing the next generation of fire effects models.