Remote Sensing of Active-Fire and Post-Fire Effects. Presentation 1-3 A Brief History of Fire-Related Remote Sensing

Transcription

1 Remote Sensing of Active-Fire and Post-Fire Effects Presentation 1-3 A Brief History of Fire-Related Remote Sensing Good Day! This lecture is entitled a brief history of fire-related remote sensing. In order to understand the potential of future remote sensing methods and datasets it is first necessary to understand where we have come from and how far we have gone. 1

2 Wildlife Management Hazard Assessment Early fire remote sensing began a few million years ago when humans first used their eyes and associated a fire with various needs and consequences. These could have included: That s a big fire lets get out of here! - So early hazard assessment; Ummm maybe we shouldn't drink this water or breath this smoke So early water and air quality assessment; or Hey those deer running from that fire may be easier to hunt So early wildlife management In essence, even in those early days we remotely sensed an object, in this case fire, evaluated how it was affecting various processes; and made a management or personnel decision. 2

3 1820s: The Photographic Age: 1826: Niepce Takes First Digital Photograph 1839: Photography begins to be widely used 1850s: First photographs taken from balloons Nadar "elevating photography to the condition of art", 1862, Honoré Daunier. In considering a brief history of fire-related remote sensing we need to step forward a few million years to the early part of the 19 th century. From the early 1820 to 1860s aerial photography from balloons starts to be widely used. 3

4 1860: Oldest Surviving Aerial Photograph As Nadar's pioneering work has been lost, the oldest surviving aerial photograph was acquired by James Wallace Black of Boston on October 13, 1860: The oldest surviving photograph that remains from this Era is from 1860 and is an oblique photograph of Boston. 4

5 1900s: The Aviation Age: 1903: Wright Brothers Invent the Airplane 1910: Wilbur Wright takes the first Aerial photographs of Italy. The 1900s ushered in the aviation age with the invention of the airplane in The Wright brothers followed up on this achievement by acquiring the first aerial photographs from an airplane in

6 Aerial and ground surveys have been used in the development of fire atlases: Fires in Northern Idaho and Western Montana: In terms of fire-related remote sensing, these early aerial photography records have been collated with ground survey records to produce what are termed fire atlases. The following description of what fire atlases are was compiled by Holden et al (2005), which alongside Morgan et al (2001) and Lentile et al (2006) are the readings for this section of this course. Holden et al (2005) writes that, Fire atlases are databases containing information of fire perimeters that are frequently used to assess fire hazard and evaluate departure from historical conditions. Fire atlases are typically constructed weeks to years after fire events using personal accounts, maps of the area burned, aerial photographs and, in recent years, satellite sensor imagery. Fire atlases do not typically include information on the internal variations within the burned area but instead provide land managers with the location and overall extent (i.e. the overall perimeter) of the area burned. 6

7 1950s FOR 435 The Space Age: 1950s: US Military invents Thermal Remote Sensing 1957: USSR Launches Sputnik 1: First Man-made Satellite in Space 1958: Invention of the Laser at Bell Labs The 1950s saw the start of the space age with the launch of Sputnik 1 in

8 1960 First ever satellite image of the Earth was taken by TIROS: TIROS = Television Infrared Observation Satellite This was the start of weather satellites. In 1960 the first weather satellite called TIROS was launched. 8

9 1972 Landsat Program Begins The start of spectral remote sensing 1972 saw a major advance in the remote sensing of the Earth with the launch of the first Landsat sensor. In terms of remote sensing milestones, the Landsat program was a big one as it ushered in the utility of spectral remote sensing. Prior to Landsat most data analysis was based on expert knowledge or the analysis of patterns, while Landsat provided actual measures of reflectance in multiple bands. In terms of fire-related remote sensing, The Landsat series of satellites have and continue to this day to be widely used to evaluate the areas burned by fires and also in recent years how these fires have impacted the environment. 9

10 1980s Present day: AVHRR used to map burned areas 1999: Launch of MODIS on the Terra satellite In addition to more Landsat sensors, the 1980s to 1990s saw the development of very low spatial resolution sensors being used for the assessment of burned area at the regional to continental-scales. Sensors such as the Advanced Very High Resolution Radiometer, or AVHRR for short, which has a pixel size of about 1.1 km on each side have been used for the last 20 years to map and monitor burned areas over expansive areas like southern Africa or the boreal forests in Russia saw the launch of the Moderate resolution Imaging Spectrometer, or MODIS for short, which in recent years has been the most widely used sensor to both map burned areas, but to also as shown in this image, identify active fire pixels. 10

11 1980s: Light Detection and Ranging (Lidar) systems are used to map particulates in the atmosphere The 1980s also saw the first development of light detection and ranging or Lidar systems. These were used for identifying particulates within air like for example those produced due to fires. 11

12 1990s: Lidar is used for the assessment of canopy fuels The 1990s saw these lidar system beginning to be used to measure canopy fuel characteristics within forested environments. In its simplest form a lidar is merely a laser rangefinder that has been stuck to the belly of an airplane. These days, lidar systems provide very accurate measures of maximum tree height and are frequently used to produce very accurate digital elevation models or DEMs. For instance, Lidar-based DEMs are frequently produced to a resolution of 1m. 12

13 1999: Key and Benson develop NBR and CBI To produce remote sensing maps of the burn severity Landsat 7:5:4 NBR One of the most important fire-related remote sensing events was in 1999, when Carl Key of the USGS and Nate Benson of the National Park Service first proposed the new remote sensing method called the Differenced Normalized Burn Ratio or delta NBR. The delta NBR method is used to classify a Landsat image of a burned area into low, moderate and high burn severity. Although it has been seven years since this method was proposed, a robust validation of it is still needed. That being said it is already (without that validation) being used by multiple federal agencies as the de-facto standard method to map the burn severity of wildfires. 13

14 1970s: Missoula Fire Lab conducts FRP research for DoD 1998: Yoram Kaufman publishes 1 st MODIS FRP paper Lets now briefly mention some advances in the remote sensing of active fires. Following early work by Wilson et al in the 1970s where R&D was performed by the Fire Lab in Missoula, remote sensing in the middle infrared (MIR: 3.9 μm) and thermal infrared (TIR: 11 μm) has allowed the rate of energy radiated by fires to be quantified (Kaufman et al. 1998; Wooster et al. 2003; Butler et al 2004; Ichoku and Kaufman 2005; Smith and Wooster 2005). This quantity, termed the fire radiative power (FRP) has been shown to be directly proportional to the rate of fuel (and carbon) volatilized (Kaufman et al 1998; Wooster 2002; Wooster et al 2005). A couple of sensor exist that use this information. Most notably are MODIS and the WASP sensor, which is operated by the Rochester Institute of Technology. Ever since Yoram Kaufman published a paper on the potential of MODIS data in 1998, MODIS data has been widely used to evaluate active fire pixels. The image here is an instantaneous measure of the power being radiated by a wildfire ad acquired by the WASP 14

15 Day of Burn The MODIS and other sensors are widely used to map the location and timing of fires at continental scales. This image, which was produced by researchers at Kings College London, shows over the span of 12 days the location and timing of fires in central Africa. 15