LANDFIRE’s (LF) Annual Disturbance products provide temporal and spatial information related to landscape change. Annual Disturbance depicts areas of 4.5 hectares (11 acres) or larger that have experienced a natural or anthropogenic landscape change (or treatment) within a given year. For the creation of the Annual Disturbance product, information sources include national fire mapping programs such as Monitoring Trends in Burn Severity (MTBS), Burned Area Reflectance Classification (BARC) and Rapid Assessment of Vegetation Condition after Wildfire (RAVG), 18 types of agency-contributed "event" perimeters (see LF Public Events Geodatabase), and remotely sensed Landsat imagery. To create the LF Annual Disturbance products, individual Landsat scenes are stacked and made into composites representing the 50th percentile of all stacked pixels (band-by-band) to reduce data gaps caused by clouds or other anomalies. Composite imagery from the specified mapping year, the two prior years, and the following year serve as the base data from which change products such as the Normalized Differenced Vegetation Index (dNDVI), the Normalized Burn Ratio (dNBR), and the Multi-Index Integrated Change Algorithm (MIICA) (Jin et al. 2013) are derived. Image analysts collectively use these datasets (separately or in combination) to isolate the true change from false change (commission errors). False changes can be attributed to many anomalies but are mostly caused by differences in annual or seasonal phenology, and/or artifacts in the image composites. Fire-caused disturbances sourced from MTBS may contain data gaps where clouds obscure the full burn scar from being mapped. Models trained from pre-fire and post-fire Landsat data are used to fill these gaps. The result is gap-free continuous severity and extent information for all MTBS fire disturbances. MTBS pixels derived from gap filling techniques, such as modeling, are noted as such in the Annual Disturbance attribute table. Smaller fires that do not meet the size criteria set forth by MTBS may be attributed using Burned Area (BA), informed from Landsat Level-3 science products and only available in the lower 48 states. Causality and severity information assigned to a disturbance are prioritized by source, with the highest priorities reserved for fire mapping programs (MTBS, BARC, and RAVG) followed by user-contributed events contained in the LF Events Geodatabase, and lastly, Landsat image-based change.

LANDFIRE’s (LF) Annual Disturbance products provide temporal and spatial information related to landscape change. Annual Disturbance depicts areas of 4.5 hectares (11 acres) or larger that have experienced a natural or anthropogenic landscape change (or treatment) within a given year. For the creation of the Annual Disturbance product, information sources include national fire mapping programs such as Monitoring Trends in Burn Severity (MTBS), Burned Area Reflectance Classification (BARC) and Rapid Assessment of Vegetation Condition after Wildfire (RAVG), 18 types of agency-contributed "event" perimeters (see LF Public Events Geodatabase), and remotely sensed Landsat imagery. To create the LF Annual Disturbance products, individual Landsat scenes are stacked and made into composites representing the 50th percentile of all stacked pixels (band-by-band) to reduce data gaps caused by clouds or other anomalies. Composite imagery from the specified mapping year, the two prior years, and the following year serve as the base data from which change products such as the Normalized Differenced Vegetation Index (dNDVI), the Normalized Burn Ratio (dNBR), and the Multi-Index Integrated Change Algorithm (MIICA) (Jin et al. 2013) are derived. Image analysts collectively use these datasets (separately or in combination) to isolate the true change from false change (commission errors). False changes can be attributed to many anomalies but are mostly caused by differences in annual or seasonal phenology, and/or artifacts in the image composites. Fire-caused disturbances sourced from MTBS may contain data gaps where clouds obscure the full burn scar from being mapped. Models trained from pre-fire and post-fire Landsat data are used to fill these gaps. The result is gap-free continuous severity and extent information for all MTBS fire disturbances. MTBS pixels derived from gap filling techniques, such as modeling, are noted as such in the Annual Disturbance attribute table. Smaller fires that do not meet the size criteria set forth by MTBS may be attributed using Burned Area (BA), informed from Landsat Level-3 science products and only available in the lower 48 states. Causality and severity information assigned to a disturbance are prioritized by source, with the highest priorities reserved for fire mapping programs (MTBS, BARC, and RAVG) followed by user-contributed events contained in the LF Events Geodatabase, and lastly, Landsat image-based change.

LANDFIRE’s (LF) Annual Disturbance products provide temporal and spatial information related to landscape change. Annual Disturbance depicts areas of 4.5 hectares (11 acres) or larger that have experienced a natural or anthropogenic landscape change (or treatment) within a given year. For the creation of the Annual Disturbance product, information sources include national fire mapping programs such as Monitoring Trends in Burn Severity (MTBS), Burned Area Reflectance Classification (BARC) and Rapid Assessment of Vegetation Condition after Wildfire (RAVG), 18 types of agency-contributed "event" perimeters (see LF Public Events Geodatabase), and remotely sensed Landsat imagery. To create the LF Annual Disturbance products, individual Landsat scenes are stacked and made into composites representing the 50th percentile of all stacked pixels (band-by-band) to reduce data gaps caused by clouds or other anomalies. Composite imagery from the specified mapping year, the two prior years, and the following year serve as the base data from which change products such as the Normalized Differenced Vegetation Index (dNDVI), the Normalized Burn Ratio (dNBR), and the Multi-Index Integrated Change Algorithm (MIICA) (Jin et al. 2013) are derived. Image analysts collectively use these datasets (separately or in combination) to isolate the true change from false change (commission errors). False changes can be attributed to many anomalies but are mostly caused by differences in annual or seasonal phenology, and/or artifacts in the image composites. Fire-caused disturbances sourced from MTBS may contain data gaps where clouds obscure the full burn scar from being mapped. Models trained from pre-fire and post-fire Landsat data are used to fill these gaps. The result is gap-free continuous severity and extent information for all MTBS fire disturbances. MTBS pixels derived from gap filling techniques, such as modeling, are noted as such in the Annual Disturbance attribute table. Smaller fires that do not meet the size criteria set forth by MTBS may be attributed using Burned Area (BA), informed from Landsat Level-3 science products and only available in the lower 48 states. Causality and severity information assigned to a disturbance are prioritized by source, with the highest priorities reserved for fire mapping programs (MTBS, BARC, and RAVG) followed by user-contributed events contained in the LF Events Geodatabase, and lastly, Landsat image-based change.

LANDFIRE's Annual Disturbance products track how landscapes change across space and time on an annual basis. The Annual Disturbance (Dist) product identifies satellite-detected areas larger than 4.5 hectares (11 acres) that underwent natural or human-caused changes within a specific year (for Dist23, October 1, 2022 – September 30, 2023), or represent fire activity/field treatments as small as 80 square meters. While creating the Annual Disturbance product a variety of data sources are leveraged. 1) National fire mapping programs: This includes information from Monitoring Trends in Burn Severity (MTBS), Burned Area Reflectance Classification (BARC), and Rapid Assessment of Vegetation Condition after Wildfire (RAVG), which offer severity information for fire-caused disturbances. 2) Agency-reported events: There are 18 designated classes for contributed polygon "Event" types such as disease, insects, development, harvest, etc. that are reported by government agencies for inclusion into the disturbance product. 3) Remotely sensed imagery: Harmonized Landsat Sentinel (HLS) satellite images offer a comprehensive-uninterrupted view of the landscape covering all lands, public and private, to fill in the gaps inherent in the previous data sources. These data are reviewed and edited by a team of image analysts to ensure and maintain high quality standards. To create the LF Annual Disturbance product, individual Landsat scenes are stacked and made into composites representing the 15th, 50th, and 90th percentiles of all stacked pixels (band-by-band) to reduce data gaps caused by clouds or other anomalies. Composite imagery from the specified mapping year and the two prior years serves as the base data from which change products such as the Normalized Differenced Vegetation Index (dNDVI), the Normalized Burn Ratio (dNBR), and the Multi-Index Integrated Change Algorithm (MIICA) (Jin et al. 2013) are derived. Image analysts collectively use these datasets (separately or in combination) to isolate the true change from false change (commission errors). False changes can be attributed to many anomalies but are most commonly caused by differences in annual or seasonal phenology, artifacts in the image composites, or difficult to map classes such as wetlands and grasses. Fire-caused disturbances sourced from MTBS may contain data gaps where clouds obscure the full burn scar from being mapped. Models trained from pre-fire and post-fire Landsat data are used to fill these gaps. The result is gap-free continuous severity and extent information for all MTBS fire disturbances. MTBS pixels derived from modeling are noted as such in the Annual Disturbance attribute table. Smaller fires that do not meet the size criteria set forth by MTBS may be attributed as fire by using Burned Area (BA) Level-3 science products derived from Landsat 8 and 9. BA data is only available in the lower 48 states (CONUS). Causality information assigned to annual disturbance products are prioritized by source, with the highest priorities reserved for fire mapping program data (MTBS, BARC, and RAVG) followed by user-contributed events contained in the LF Events Geodatabase, and lastly, satellite image-based change. Severity is assigned directly from fire program data. For events and satellite-detected change, severity is derived from pre- and post-burn standard deviation values of the differenced Normalized Burn Ratio (dNBR). When mapping the LF Annual Disturbance product, the start date is utilized for disturbances from fire program data whereas all other disturbances utilize the end date.

LANDFIRE's Annual Disturbance products track how landscapes change across space and time on an annual basis. The Annual Disturbance (Dist) product identifies satellite-detected areas larger than 4.5 hectares (11 acres) that underwent natural or human-caused changes within a specific year (for Dist23, October 1, 2022 – September 30, 2023), or represent fire activity/field treatments as small as 80 square meters. While creating the Annual Disturbance product a variety of data sources are leveraged. 1) National fire mapping programs: This includes information from Monitoring Trends in Burn Severity (MTBS), Burned Area Reflectance Classification (BARC), and Rapid Assessment of Vegetation Condition after Wildfire (RAVG), which offer severity information for fire-caused disturbances. 2) Agency-reported events: There are 18 designated classes for contributed polygon "Event" types such as disease, insects, development, harvest, etc. that are reported by government agencies for inclusion into the disturbance product. 3) Remotely sensed imagery: Harmonized Landsat Sentinel (HLS) satellite images offer a comprehensive-uninterrupted view of the landscape covering all lands, public and private, to fill in the gaps inherent in the previous data sources. These data are reviewed and edited by a team of image analysts to ensure and maintain high quality standards. To create the LF Annual Disturbance product, individual Landsat scenes are stacked and made into composites representing the 15th, 50th, and 90th percentiles of all stacked pixels (band-by-band) to reduce data gaps caused by clouds or other anomalies. Composite imagery from the specified mapping year and the two prior years serves as the base data from which change products such as the Normalized Differenced Vegetation Index (dNDVI), the Normalized Burn Ratio (dNBR), and the Multi-Index Integrated Change Algorithm (MIICA) (Jin et al. 2013) are derived. Image analysts collectively use these datasets (separately or in combination) to isolate the true change from false change (commission errors). False changes can be attributed to many anomalies but are most commonly caused by differences in annual or seasonal phenology, artifacts in the image composites, or difficult to map classes such as wetlands and grasses. Fire-caused disturbances sourced from MTBS may contain data gaps where clouds obscure the full burn scar from being mapped. Models trained from pre-fire and post-fire Landsat data are used to fill these gaps. The result is gap-free continuous severity and extent information for all MTBS fire disturbances. MTBS pixels derived from modeling are noted as such in the Annual Disturbance attribute table. Smaller fires that do not meet the size criteria set forth by MTBS may be attributed as fire by using Burned Area (BA) Level-3 science products derived from Landsat 8 and 9. BA data is only available in the lower 48 states (CONUS). Causality information assigned to annual disturbance products are prioritized by source, with the highest priorities reserved for fire mapping program data (MTBS, BARC, and RAVG) followed by user-contributed events contained in the LF Events Geodatabase, and lastly, satellite image-based change. Severity is assigned directly from fire program data. For events and satellite-detected change, severity is derived from pre- and post-burn standard deviation values of the differenced Normalized Burn Ratio (dNBR). When mapping the LF Annual Disturbance product, the start date is utilized for disturbances from fire program data whereas all other disturbances utilize the end date.

The LANDFIRE Preliminary Annual Disturbance (PDist)23 product is a new product introduced with the LF 2023 Update (LF 2023). While LDist23 is a first "draft", PDist23 is akin to a second “draft” of the LANDFIRE Annual Disturbance product and includes disturbance events captured through October 31, 2023. PDist23 is releasing for these extents throughout the calendar year 2024: Conterminous United States (CONUS), Alaska (AK), Hawaii (HI), and Puerto Rico and the US Virgin Islands (PRVI). See the LF 2023 page for more information as this page will be added to as more details are available. https://landfire.gov/data/lf2023 Preliminary Annual Disturbance (PDist23) was created with Landsat imagery, submitted disturbance events from the data call, and online fire program data collected on or before October 31, 2023. Unlike the Limited Annual Disturbance (LDist23) product released earlier in calendar year 2024, PDist23 includes remotely sensed severity for submitted events and unknown disturbances where change was detected.

The LANDFIRE Preliminary Annual Disturbance (PDist)23 product is a new product introduced with the LF 2023 Update (LF 2023). While LDist23 is a first "draft", PDist23 is akin to a second “draft” of the LANDFIRE Annual Disturbance product and includes disturbance events captured through October 31, 2023. PDist23 is releasing for these extents throughout the calendar year 2024: Conterminous United States (CONUS), Alaska (AK), Hawaii (HI), and Puerto Rico and the US Virgin Islands (PRVI). See the LF 2023 page for more information as this page will be added to as more details are available. https://landfire.gov/data/lf2023 Preliminary Annual Disturbance (PDist23) was created with Landsat imagery, submitted disturbance events from the data call, and online fire program data collected on or before October 31, 2023. Unlike the Limited Annual Disturbance (LDist23) product released earlier in calendar year 2024, PDist23 includes remotely sensed severity for submitted events and unknown disturbances where change was detected.

The LANDFIRE Limited Disturbance (LDist)23 product is a new product introduced with the LF 2023 Update (LF 2023). LDist23 is an early "draft" of the LANDFIRE Annual Disturbance product and includes disturbance events captured through October 31, 2023. LDist23 is the first of three Annual Disturbance products releasing in the LF 2023 Update. LDist23 releases in January 2024, then Preliminary Disturbance (PDist)23 will release mid-year 2024. PDist23 will be the second "draft" of Annual Disturbance for the LF 2023 update. Finally, in the fall of 2024 the final "draft" of Annual Disturbance (Dist23) will be released.

The LANDFIRE Limited Disturbance (LDist)23 product is a new product introduced with the LF 2023 Update (LF 2023). LDist23 is an early "draft" of the LANDFIRE Annual Disturbance product and includes disturbance events captured through October 31, 2023. LDist23 is the first of three Annual Disturbance products releasing in the LF 2023 Update. LDist23 releases in January 2024, then Preliminary Disturbance (PDist)23 will release mid-year 2024. PDist23 will be the second "draft" of Annual Disturbance for the LF 2023 update. Finally, in the fall of 2024 the final "draft" of Annual Disturbance (Dist23) will be released.

LANDFIRE's (LF) Annual Disturbance (Dist) product provides temporal and spatial information related to landscape change. Dist depicts areas that have experienced a disturbance within a given year of 4.5 hectares (11 acres) or larger, along with cause and severity. Information sources include national fire mapping programs such as Monitoring Trends in Burn Severity (MTBS), Burned Area Reflectance Classification (BARC), and Rapid Assessment of Vegetation Condition after Wildfire (RAVG), local user/agency contributed data (LF Events Geodatabase), and remotely sensed Landsat imagery. Composite Landsat image pairs from the current year, prior year, and following year are spectrally compared to determine where change occurred and its corresponding severity. Additionally, vegetation indices (Normalized Differenced Vegetation Index [NDVI] and Normalized Burn Ratio [NBR]) serve as inputs into the Multi-Index Integrated Change Algorithm (MIICA) (Jin et al. 2013); MIICA outputs and differenced products (e.g., dNDVI and dNBR) are used to locate change. Predictive modeling based on the previous 10 years of disturbance data provides an additional dataset useful for locating disturbance. Image analysts use the aforementioned datasets separately or in combination to isolate true change from false change (e.g., change caused by stark differences in phenology rather than a true disturbance event). The accuracy of the final product is often related to the quality of the Landsat image composite. Areas with persistent cloud cover are particularly challenging (e.g., the northeast US). Fire caused disturbances sourced from MTBS may contain data gaps where clouds, smoke, water or Landsat7 SLC-off stripes exist. Models trained from pre-fire and post-fire Landsat data are used to fill the gaps. The result is continuous severity and extent information for all MTBS fire disturbances. MTBS pixels derived from gap filling techniques, such as modeling, are noted as such in their corresponding attribute table. Smaller fires that do not meet the size criteria set forth by MTBS) may be attributed as a Burned Area Essential Climate Variable (BAECV), which are only produced for the lower 48 states. Causality and severity information assigned to a disturbance are prioritized by source, with the highest priorities reserved for fire mapping programs (MTBS, BARC and RAVG) followed by user-contributed events contained in the LF Events Geodatabase, and lastly, Landsat image based change.