Camp Swift Fire Experiment 2014: Active fire unmanned aerial vehicle (UAV) infrared aerial imagery and video

Metadata:

Data_Quality_Information:

Attribute_Accuracy:

Attribute_Accuracy_Report:

The time stamp for every georeferenced image is estimated based on recorded times for a subset of screen shots. The time is expected to be within one or two seconds. The digital values of the imagery are not calibrated. Changes in pixel values qualitatively represent heat sources observed in the imagery. Other attributes in the attribute table associated with the Mosaic dataset are automatically populated by the ArcGIS™ software.

Logical_Consistency_Report:

The attribute table for the georeferenced images contains a time stamp for every image georeferenced.

Completeness_Report:

There were only time estimates for the IR video for burn block 1 and burn block 3, which were used for estimating times for the EO screen shots exported at one-second intervals. Burn block 2 had no explicit times associated with any of the EO or IR video and the recorded ignition time was used to locate in time the first screen shots containing active-fire information. Times recorded in this dataset for IR images might require adjustment to be integrated with other Camp Swift data. The spatial extent covered by the active-fire IR imagery is coincident with the burn block ignited. There are some video screen shots that did not have adequate ground control and consequently were not georeferenced. The estimated transmission rate over the 2.4 Gigahertz (Ghz) wireless video transmission was never verified.

Only the video from burn block 1 was retained, the IR videos from burn block 2 and burn block 3 were never provided, only screen shots from the video.

Positional_Accuracy:

Horizontal_Positional_Accuracy:

Horizontal_Positional_Accuracy_Report:

For each video screen shot georeferenced, a root mean square error (RMSE) of less than 1.5 m was attempted to be achieved. At least six ground control points were used to achieve this RMSE. All ground control points were on the edge of the burn blocks and there were not any ground control within the burn blocks or that extended beyond the fuel breaks. Additionally, this imagery was collected at an oblique angle. The sensor moved in a circular orbit around the burn block and as such the distortion caused by the oblique angle was continually changing and not quantified in this document. Other issues causing distortion include pixel smear caused by the forward motion of the UAV and consequently the sensor. Pixel smear from the IR imagery can be significant and many IR georeferenced images appear to suffer from pixel smear caused by the forward motion of the sensor and the time for the IR sensor to integrate the data. This in some cases can be seen in images of a later time being before, in space, images of an earlier time representing the fire line. Furthermore, the IR imagery suffered from saturation and blooming that prevents quantitative measurement when imaging high temperature items as the IR sensor is designed to image "normal" temperature objects, typically representative of flora and fauna. All sensors require some form of calibration if they are to deliver accurate measurements, small IR sensors like employed at Camp Swift have proven problematic in regards to calibration. An Edmund Optics Worldwide; 0.5 optical density, 25 millimeter (mm) diameter, IR neutral density filter was installed on the IR sensor. An associated ArcGIS Online story map (McNamara and Mell, 2018) examines potential saturation of imagery in more detail. The filter does cause a ghost image to appear representing the reversed fire line. The spatial accuracy of the fire portrayed in this dataset, therefore, is not well understood but an associated ArcGIS Online story map (McNamara and Mell, 2018) examines potential saturation of imagery in more detail.

Quantitative_Horizontal_Positional_Accuracy_Assessment:

Horizontal_Positional_Accuracy_Value: 0.5

Horizontal_Positional_Accuracy_Explanation:

Use of ground control points to examine RMSE during the georeferencing process.

Lineage:

Process_Step:

Process_Description:

A complete user manual for operation of the MLB SuperBat Unmanned Aerial Vehicle System contains specifics about deploying this system (Hennig and Ridenour-Jackson 2015). The IR sensor records information in spectral bands from 700-1500 nanometer (nm). The sensor has a 10.5 degree horizontal field of view with image resolution of 640 by 480 pixels. Data were collected using the National Television System Committee (NTSC) standard at 29.97 FPS, real-time video feed. An Edmund Optics Worldwide; 0.5 optical density, 25 mm diameter, IR neutral density filter was installed on the IR sensor to test the ability to attenuate the signal and reduce saturation, blooming and radiometric accuracy problems. Times recorded in this dataset for IR images might require adjustment to be integrated with other Camp Swift data. The sensor was attached to the MLB Company SuperBat III unmanned aerial vehicle (UAV). The SuperBat had a wing span of 2.59 meters (m), a length of 1.5 m. The gross takeoff weight with payload was 15.88 kilograms (kg). The propulsion consisted of a 26 cubic centimeter (cc) internal combustion engine. The Fuel on Board was 3.13 kg of a gas-oil mixture. The SuperBat has a flight duration of 10 hours or 400 miles but there were no documented flights for flight duration or distance due to restrictions of the certification of wavier or authorization (COA). The SuperBat is launched with a catapult system and lands with a conventional skid landing. The system has a parachute for failover recovery.

To provide for a sufficient oblique view, the aircraft flight path was set as an expanded loiter, radius with the center point of the loiter placed in the center, or near to center of each burn block. The mission altitude was assessed to allow for complete view of the burn block from all angles as the aircraft traveled through its orbit. The only limitations to flight planning for FMV acquisition where the boundaries of the TFR that was established for the event and the reduced field of view of the IR sensor on the TASE DUO deployed video payload. The TFR boundaries provided suitable operational space to perform the tasks as required. The video had a payload rated for 29.97 frames per second as the data transmission rate between the UAV sensors and the ground operations and equipment.

Active-fire video was collected with an attempt to maintain a full field of view of the entire burn block, which was accomplished through adjustment of the orbit and/or increasing the altitude of the aircraft. The altitude of the UAV during acquisition was approximately 183 m. The orbit diameter was approximately 750 m. Still screen captures of the video were exported at one-second intervals and these still images were manually georeferenced to produce a set of images where co-location in space and time with other Camp Swift data can be assessed. The focus of the active-fire data collection was on infra-red (IR) video. The differences in focal length between the EO and IR video cameras required some adjustment to maintain all targets in the field of view.

The geographic extent capable of imaging was limited by the Temporary Flight Restriction (TFR) established by the Federal Aviation Administration (FAA). The TFR was for a radius of 3.6 kilometers (km) of the Camp Swift research burns at or below 610 meters (m) above ground level. The TFR ensured deconfliction with civilian air traffic. Air operations at this location were permitted under the amended COA 2013-CSA-15-COA. Coordination with the Base Commander ensured deconfliction with military air traffic. Airstrip requirements are aircraft specific, in this case an airstrip of 100 m long by 50 m wide met the operational needs of the UAV system.

Hennig, Mike; Ridenour-Jackson, Karen. 2015. NIST WUI Project - MLB SUPERBAT Unmanned Aerial Vehicle System User Manual. NIST GCR 15-982. U.S. Department of Commerce, National Institute of Standards and Technology. https://doi.org/10.6028/NIST.GCR.15-982

Process_Date: 20140114

Process_Contact:

Contact_Information:

Contact_Organization_Primary:

Contact_Organization: Rp Flight Systems Inc

Contact_Person: Gene Robinson

Contact_Position: Pilot

Contact_Address:

Address_Type: mailing and physical

City: Wimberley

State_or_Province: Texas

Postal_Code: 78676

Country: USA

Contact_Voice_Telephone: 512-665-9990

Process_Step:

Process_Description:

After acquisition of the active fire IR imagery, individual min-DV tapes containing the transmitted video were transferred to MP4 digital formats using an unknown encoder. Screen shots of the video were then exported at 1-second intervals.

Process_Date: 20140301

Process_Contact:

Contact_Information:

Contact_Organization_Primary:

Contact_Organization: Colorado State University

Contact_Person: Michael Hennig

Contact_Position: Pilot

Contact_Address:

Address_Type: mailing and physical

City: Fort Collins

State_or_Province: Colorado

Postal_Code: 80523

Country: USA

Contact_Voice_Telephone: 619-573-0935

Process_Step:

Process_Description:

The video IR screen shots were georeferenced in ArcGIS using ground control. Typically six ground control points were used in the georeferencing process with a projective transformation used to help account for the oblique angle. A root mean square error of less than 0.5 meters was attempted to be acheved but this was not possible in all cases. The specific ground control used for georeferencing each image can be found in \Supplements\GroundControlPoints.txt.

After creating the individual georeferenced active-fire IR images, they were loaded to this file geodatabase. A mosaic dataset was created in this file geodatabase and the individual georeferenced rasters were loaded to this dataset. The images were aligned with captured IR video frames during the fire containing times. The best fit IR frame to the IR georeferenced frame was used as an established time and each subsequent image had one-second added to the time and proceeding images had one-second subtracted from the time. This process considered those frames that were not georeferenced due to signal drop out or other noise in the data. Only burn block 1 and burn block 3 had times for IR frames. There was no time information provided for burn block 2. The first active fire image from the IR video, consequently, had the time set to the ignition time recorded in the field. Note that in burn block 1 and burn block 3, the ignition time as recorded in the UAV acquired video were different than the time recorded on the ground. The ground recorded ignition time is found Camp Swift Fire Experiment 2014: Ignition Points (McNamara, 2018).

McNamara, Derek J.; William, Mell E. 2018. Camp Swift Fire Experiment 2014: Ignition points. Fort Collins, CO: Forest Service Research Data Archive. https://doi.org/10.2737/RDS-2018-0045

Process_Date: 20141101

Process_Contact:

Contact_Information:

Contact_Organization_Primary:

Contact_Organization: Geospatial Measurement Solutions, LLC

Contact_Person: Derek McNamara

Contact_Position: GIS Analyst

Contact_Address:

Address_Type: physical

Address: 1841 4th Street

City: Hood River

State_or_Province: OR

Postal_Code: 97031

Country: USA

Contact_Voice_Telephone: 208-818-2152

Contact_Electronic_Mail_Address: dmgeo@gmsgis.com

Spatial_Data_Organization_Information:

Direct_Spatial_Reference_Method: Raster

Raster_Object_Information:

Raster_Object_Type: Grid Cell

Spatial_Reference_Information:

Horizontal_Coordinate_System_Definition:

Planar:

Map_Projection:

Map_Projection_Name: NAD 1983 UTM Zone 14N

Transverse_Mercator:

Scale_Factor_at_Central_Meridian: 0.9996

Longitude_of_Central_Meridian: -99.0

Latitude_of_Projection_Origin: 0.0

False_Easting: 500000.0

False_Northing: 0.0

Planar_Coordinate_Information:

Planar_Coordinate_Encoding_Method: coordinate pair

Coordinate_Representation:

Abscissa_Resolution: 0.000000002220024164500956

Ordinate_Resolution: 0.000000002220024164500956

Planar_Distance_Units: meter

Geodetic_Model:

Horizontal_Datum_Name: D North American 1983

Ellipsoid_Name: GRS 1980

Semi-major_Axis: 6378137.0

Denominator_of_Flattening_Ratio: 298.257222101

Entity_and_Attribute_Information:

Detailed_Description:

Entity_Type:

Entity_Type_Label: BurnBlock#IR

Entity_Type_Definition:

Group of raster datasets portraying active-fire imagery and bounding polygons.

Entity_Type_Definition_Source:

Geospatial Measurement Solutions, LLC

Attribute:

Attribute_Label: OBJECTID

Attribute_Definition:

Internal feature number.

Attribute_Definition_Source:

ESRI

Attribute_Domain_Values:

Unrepresentable_Domain:

Sequential unique whole numbers that are automatically generated.

Attribute:

Attribute_Label: Shape

Attribute_Definition:

Feature geometery for index polygons.

Attribute_Definition_Source:

Esri

Attribute_Domain_Values:

Unrepresentable_Domain:

Coordinates defining the features.

Attribute:

Attribute_Label: Raster

Attribute_Definition:

Binary large object containing the individual active fire image.

Attribute_Definition_Source:

ESRI

Attribute_Domain_Values:

Unrepresentable_Domain:

Binary large object.

Attribute:

Attribute_Label: Name

Attribute_Definition:

The name of the respective active-fire IR georeferenced video frame.

Attribute_Definition_Source:

Geospatial Measurement Solutions, LLC

Attribute_Domain_Values:

Unrepresentable_Domain:

The name of the raster dataset.

Attribute:

Attribute_Label: MinPS

Attribute_Definition:

The minimum pixel size, computed from the LowPS and HighPS attribute values and depends on the overlapping raster datasets.

Attribute_Definition_Source:

ESRI

Attribute_Domain_Values:

Range_Domain:

Range_Domain_Minimum: 0

Range_Domain_Maximum: 1.594749

Attribute:

Attribute_Label: maxPS

Attribute_Definition:

The maximum pixel size, computed from the LowPS and HighPS attribute values and depends on the overlapping raster datasets.

Attribute_Definition_Source:

ESRI

Attribute_Domain_Values:

Range_Domain:

Range_Domain_Minimum: 0.465324

Range_Domain_Maximum: 79.73745

Attribute:

Attribute_Label: LowPS

Attribute_Definition:

The lowest range of pixel sizes that this raster dataset contains.

Attribute_Definition_Source:

ESRI

Attribute_Domain_Values:

Range_Domain:

Range_Domain_Minimum: 0.253019

Range_Domain_Maximum: 1.594749

Attribute:

Attribute_Label: HighPS

Attribute_Definition:

The highest range of pixel sizes that this raster dataset contains.

Attribute_Definition_Source:

ESRI

Attribute_Domain_Values:

Range_Domain:

Range_Domain_Minimum: 0.253019

Range_Domain_Maximum: 1.594749

Attribute:

Attribute_Label: Category

Attribute_Definition:

Used to quickly identify the type of dataset and its status within the mosaic dataset.

Attribute_Definition_Source:

ESRI

Attribute_Domain_Values:

Unrepresentable_Domain:

1—Primary (base) data, 2—Overview

Attribute:

Attribute_Label: Tag

Attribute_Definition:

Used for identifying the raster dataset that will participate in functions defined in a function template. In most cases, the value is defined by the raster type used for creating the dataset entry within the table.

Attribute_Definition_Source:

ESRI

Attribute_Domain_Values:

Unrepresentable_Domain:

Dataset

Attribute:

Attribute_Label: GroupName

Attribute_Definition:

A name given to one or more rows in the table that belong together.

Attribute_Definition_Source:

ESRI

Attribute_Domain_Values:

Unrepresentable_Domain:

Not populated in this dataset.

Attribute:

Attribute_Label: ProductName

Attribute_Definition:

This is derived from either the name or product type defined in the raster type.

Attribute_Definition_Source:

ESRI

Attribute_Domain_Values:

Unrepresentable_Domain:

Not populated in this dataset.

Attribute:

Attribute_Label: CenterX

Attribute_Definition:

The x-coordinate identifying the centroid or nadir of the raster dataset.

Attribute_Definition_Source:

ESRI

Attribute_Domain_Values:

Range_Domain:

Range_Domain_Minimum: 666150.505802

Range_Domain_Maximum: 666399.448

Attribute:

Attribute_Label: CenterY

Attribute_Definition:

The y-coordinate identifying the centroid or nadir of the raster dataset.

Attribute_Definition_Source:

ESRI

Attribute_Domain_Values:

Range_Domain:

Range_Domain_Minimum: 3347195.7864

Range_Domain_Maximum: 3347336.09105

Attribute:

Attribute_Label: Zorder

Attribute_Definition:

Controls the ordering used in determining how the rasters will be mosaicked together when using the Closest To Center, North-West, By Attribute, Closest To Nadir, or Closest To Viewpoint mosaic methods.

Attribute_Definition_Source:

ESRI

Attribute_Domain_Values:

Unrepresentable_Domain:

Not populated in this dataset.

Attribute:

Attribute_Label: TypeID

Attribute_Definition:

Raster type ID

Attribute_Definition_Source:

ESRI

Attribute_Domain_Values:

Range_Domain:

Range_Domain_Minimum: 1

Range_Domain_Maximum: 415

Attribute:

Attribute_Label: ItemTS

Attribute_Definition:

Unknown

Attribute_Definition_Source:

ESRI

Attribute:

Attribute_Label: UriHash

Attribute_Definition:

Unique ID field.

Attribute_Definition_Source:

ESRI

Attribute_Domain_Values:

Unrepresentable_Domain:

Global unique identifiers.

Attribute:

Attribute_Label: Uri

Attribute_Definition:

Binary large object.

Attribute_Definition_Source:

ESRI

Attribute_Domain_Values:

Unrepresentable_Domain:

Binary large object.

Attribute:

Attribute_Label: Shape_Length

Attribute_Definition:

The length of the respective image bounding box polygon in meters.

Attribute_Definition_Source:

ESRI

Attribute:

Attribute_Label: Shape_Area

Attribute_Definition:

The area of the respective image bounding box polygon in meters.

Attribute_Definition_Source:

ESRI

Attribute:

Attribute_Label: Thumbnail

Attribute_Definition:

A thumbnail of the respective image.

Attribute_Definition_Source:

ESRI

Attribute_Domain_Values:

Unrepresentable_Domain:

Binary large object containing a thumbnail view of the raster dataset.

Attribute:

Attribute_Label: FullDate

Attribute_Definition:

The full date and time, to the nearest second, of the respective active-fire georeferenced video frame.

Attribute_Definition_Source:

Geospatial Measurement Solutions, LLC

Attribute_Domain_Values:

Unrepresentable_Domain:

Date and time field.

Overview_Description:

Entity_and_Attribute_Overview:

DATA

\Data\GeoReferencedFrames\BurnBlock#IR.gdb: Environmental Systems Research Institute (ESRI) File Geodatabase (FGDB) containing individual georeferenced video frames record by the IR sensor. Also contained within the FGDB is a mosaic dataset containing all the individual georeferenced video frames, with associated time estimates in coordinated universal time (UTC). (See the entity and attribute detail section above for a complete description of the data available in this file.)

\Data\Videos\2014_01_15 3195TASE_IR tape1.mp4: MPEG4 video recorded by the IR sensor portraying burn block 1 active-fire conditions as recorded on January 15, 2014.

Entity_and_Attribute_Detail_Citation:

None provided

Overview_Description:

Entity_and_Attribute_Overview:

SUPPLEMENTAL FILES

NOTE: in the attached files, plot = burn block

Camp Swift Overview

\Supplements\CampSwiftFireExperiment2014_BurnBlockDesign.pdf: Adobe Acrobat PDF/a file containing a description of the 2014 Camp Swift Study Sites and processes used to create the burn block design.

\Supplements\CampSwiftFireExperiment2014_Project_Overview.pdf: Adobe Acrobat PDF/a document providing an overview of the Camp Swift effort.

\Supplements\GroundControlPoints.txt: ASCII text file containing the latitude and longitude (in decimal degrees) of ground control points used for georeferencing. These points represent the center of the fire shelters as recorded with a GEOXH GeoExplorer 6000 series Trimble - differential Global Positioning System (GPS) unit. Post-processing of GPS collected data was done in Pathfinder Office 5.4. A minimum of four satellites were required for recording GPS points found in this dataset. The Global Navigation Satellite System (GNSS) Smart Setting for position dilution of precision (PDOP), signal to noise ratio (SNR) and estimate of horizontal error (EHE) on the Trimble unit was selected for this site, requiring no adjustments of settings within the study area. Wide Area Augmentation Systems (WAAS) along with post-processing enhanced the precision of the data derived.


Instrument Locations & Specifications

\Supplements\InstrumentIgnitionLocations.pdf: Geospatial PDF file containing a map showing instrument locations.

\Supplements\InstrumentSpecifications.pdf: Adobe Acrobat PDF/a file containing a list of the instruments used in this study along with their specifications, such as the type, accuracy, measurements, etc.

Entity_and_Attribute_Detail_Citation:

None provided

Metadata_Reference_Information:

Metadata_Date: 20180927

Metadata_Contact:

Contact_Information:

Contact_Organization_Primary:

Contact_Organization: Geospatial Measurement Solutions, LLC

Contact_Person: Derek McNamara

Contact_Position: GIS Analyst

Contact_Address:

Address_Type: physical

Address: 1841 4th Street

City: Hood River

State_or_Province: OR

Postal_Code: 97031

Country: USA

Contact_Voice_Telephone: 208-818-2152

Contact_Electronic_Mail_Address: dmgeo@gmsgis.com

Metadata_Standard_Name: FGDC Content Standard for Digital Geospatial Metadata

Metadata_Standard_Version: FGDC-STD-001-1998