Geology topics

Joshua J. Picotte

Publications and source records attributed to Joshua J. Picotte.

24 records · Page 2Linked to original sources

USGS2016

An optimal sample data usage strategy to minimize overfitting and underfitting effects in regression tree models based on remotely-sensed data

Regression tree models have been widely used for remote sensing-based ecosystem mapping. Improper use of the sample data (model training and testing data) may cause overfitting and underfitting effects in the model. The goal of this study is to develop an optimal sampling data usage strategy for any dataset and identify an appropriate number of rules in the regression tree model that will improve its accuracy and robustness. Landsat 8 data and Moderate-Resolution Imaging Spectroradiometer-scaled Normalized Difference Vegetation Index (NDVI) were used to develop regression tree models. A Python procedure was designed to generate random replications of model parameter options across a range of model development data sizes and rule number constraints. The mean absolute difference (MAD) between the predicted and actual NDVI (scaled NDVI, value from 0–200) and its variability across the different randomized replications were calculated to assess the accuracy and stability of the models. In our case study, a six-rule regression tree model developed from 80% of the sample data had the lowest MAD (MAD training = 2.5 and MAD testing = 2.4), which was suggested as the optimal model. This study demonstrates how the training data and rule number selections impact model accuracy and provides important guidance for future remote-sensing-based ecosystem modeling.

Remote Sensing↗

USGS2016

1984–2010 trends in fire burn severity and area for the conterminous US

Burn severity products created by the Monitoring Trends in Burn Severity (MTBS) project were used to analyse historical trends in burn severity. Using a severity metric calculated by modelling the cumulative distribution of differenced Normalized Burn Ratio (dNBR) and Relativized dNBR (RdNBR) data, we examined burn area and burn severity of 4893 historical fires (1984–2010) distributed across the conterminous US (CONUS) and mapped by MTBS. Yearly mean burn severity values (weighted by area), maximum burn severity metric values, mean area of burn, maximum burn area and total burn area were evaluated within 27 US National Vegetation Classification macrogroups. Time series assessments of burned area and severity were performed using Mann–Kendall tests. Burned area and severity varied by vegetation classification, but most vegetation groups showed no detectable change during the 1984–2010 period. Of the 27 analysed vegetation groups, trend analysis revealed burned area increased in eight, and burn severity has increased in seven. This study suggests that burned area and severity, as measured by the severity metric based on dNBR or RdNBR, have not changed substantially for most vegetation groups evaluated within CONUS.

International Journal of Wildland Fire↗

USGS2014

Utilizing multi-sensor fire detections to map fires in the United States

In 2006, the Monitoring Trends in Burn Severity (MTBS) project began a cooperative effort between the US Forest Service (USFS) and the U.S.Geological Survey (USGS) to map and assess burn severity all large fires that have occurred in the United States since 1984. Using Landsat imagery, MTBS is mandated to map wildfire and prescribed fire that meet specific size criteria: greater than 1000 acres in the west and 500 acres in the east, regardless of ownership. Relying mostly on federal and state fire occurrence records, over 15,300 individual fires have been mapped. While mapping recorded fires, an additional 2,700 “unknown” or undocumented fires were discovered and assessed. It has become apparent that there are perhaps thousands of undocumented fires in the US that are yet to be mapped. Fire occurrence records alone are inadequate if MTBS is to provide a comprehensive accounting of fire across the US. Additionally, the sheer number of fires to assess has overwhelmed current manual procedures. To address these problems, the National Aeronautics and Space Administration (NASA) Applied Sciences Program is helping to fund the efforts of the USGS and its MTBS partners (USFS, National Park Service) to develop, and implement a system to automatically identify fires using satellite data. In near real time, USGS will combine active fire satellite detections from MODIS, AVHRR and GOES satellites with Landsat acquisitions. Newly acquired Landsat imagery will be routinely scanned to identify freshly burned area pixels, derive an initial perimeter and tag the burned area with the satellite date and time of detection. Landsat imagery from the early archive will be scanned to identify undocumented fires. Additional automated fire assessment processes will be developed. The USGS will develop these processes using open source software packages in order to provide freely available tools to local land managers providing them with the capability to assess fires at the local level.

Conference Paper↗

USGS2014

Use of multi-sensor active fire detections to map fires in the United States: the future of monitoring trends in burn severity

The effort to utilize satellite-based MODIS, AVHRR, and GOES fire detections from the Hazard Monitoring System (HMS) to identify undocumented fires in Florida and improve the Monitoring Trends in Burn Severity (MTBS) mapping process has yielded promising results. This method was augmented using regression tree models to identify burned/not-burned pixels (BnB) in every Landsat scene (1984–2012) in Worldwide Referencing System 2 Path/Rows 16/40, 17/39, and 1839. The burned area delineations were combined with the HMS detections to create burned area polygons attributed with their date of fire detection. Within our study area, we processed 88,000 HMS points (2003–2012) and 1,800 Landsat scenes to identify approximately 300,000 burned area polygons. Six percent of these burned area polygons were larger than the 500-acre MTBS minimum size threshold. From this study, we conclude that the process can significantly improve understanding of fire occurrence and improve the efficiency and timeliness of assessing its impacts upon the landscape.

Conference Paper↗

USGS2013

LANDFIRE 2010 - updated data to support wildfire and ecological management

Wildfire is a global phenomenon that affects human populations and ecosystems. Wildfire effects occur at local to global scales impacting many people in different ways (Figure 1). Ecological concerns due to land use, fragmentation, and climate change impact natural resource use, allocation, and conservation. Access to consistent and current environmental data is a constant challenge, yet necessary for understanding the complexities of wildfire and ecological management. Data products and tools from the LANDFIRE Program help decision-makers to clarify problems and identify possible solutions when managing fires and natural resources. LANDFIRE supports the reduction of risk from wildfire to human lives and property, monitoring of fire danger, prediction of fire behavior on active incidents, and assessment of fire severity and impacts on natural systems [1] [2] [3]. LANDFIRE products are unique in that they are nationally consistent and provide the only complete geospatial dataset describing vegetation and wildland fuel information for the entire U.S. As such, LANDFIRE data are useful for many ecological applications [3]. For example, LANDFIRE data were recently integrated into a decision-support system for resource management and conservation decision-making along the Appalachian Trail. LANDFIRE is a joint effort between the U.S. Department of the Interior Office of Wildland Fire, U.S. Department of Agriculture Forest Service Fire & Aviation Management, and The Nature Conservancy. To date, seven versions of LANDFIRE data have been released, with each successive version improving the quality of the data, adding additional features, and/or updating the time period represented by the data. The latest version, LANDFIRE 2010 (LF 2010), released mid-2013, represents circa 2010 landscape conditions and succeeds LANDFIRE 2008 (LF 2008), which represented circa 2008 landscape conditions. LF 2010 used many of the same processes developed for the LF 2008 effort [3]. Ongoing refinement of the LANDFIRE vegetation and fuel data is necessary to improve the quality and usability of the data and to capture landscape disturbance. LANDFIRE relies on Landsat multi-spectral imagery to produce and update vegetation and fuel data. The deep Landsat archive provides data needed for vegetation classification, change analysis, and historical disturbance characterization, for which LANDFIRE has used more than 24,000 image scenes since the program’s inception. In addition, LF 2010 used airborne and spaceborne lidar, and spaceborne synthetic aperture radar (SAR) to map vegetation structure in areas where ground-based field information was lacking, including Alaska and U.S.-affiliated islands in the Caribbean and the Pacific. The mapping of insular areas is new for the 2010 data release; previous versions of LANDFIRE were limited to the conterminous U.S., Alaska, and Hawaii.

Earthzine↗

USGS2011

Timing constraints on remote sensing of wildland fire burned area in the southeastern US

Remote sensing using Landsat Thematic Mapper (TM) satellite imagery is increasingly used for mapping wildland fire burned area and burn severity, owing to its frequency of collection, relatively high resolution, and availability free of charge. However, rapid response of vegetation following fire and frequent cloud cover pose challenges to this approach in the southeastern US. We assessed these timing constraints by using a series of Landsat TM images to determine how rapidly the remotely sensed burn scar signature fades following prescribed burns in wet flatwoods and depression swamp community types in the Apalachicola National Forest, Florida, USA during 2006. We used both the Normalized Burn Ratio (NBR) of reflectance bands sensitive to vegetation and exposed soil cover, as well as the change in NBR from before to after fire (dNBR), to estimate burned area. We also determined the average and maximum amount of time following fire required to obtain a cloud-free image for burns in each month of the year, as well as the predicted effect of this time lag on percent accuracy of burn scar estimates. Using both NBR and dNBR, the detectable area decreased linearly 9% per month on average over the first four months following fire. Our findings suggest that the NBR and dNBR methods for monitoring burned area in common southeastern US vegetation community types are limited to an average of 78–90% accuracy among months of the year, with individual burns having values as low as 38%, if restricted to use of Landsat 5 TM imagery. However, the majority of burns can still be mapped at accuracies similar to those in other regions of the US, and access to additional sources of satellite imagery would improve overall accuracy.

Remote Sensing↗