Fort McMurray Wildfires and Near-Constant Contrast (NCC) Imagery

The Fort McMurray Wildfires started in the city of Fort McMurray, located in the northeastern part of Alberta, a province of Canada. The wildfires started 01 May 2016 and are still currently burning. The wildfires have burned over 1,200,000 plus acres of land and has reached into parts of western Saskatchewan. Over 2,400 plus homes and businesses were lost within the Fort McMurray area (The Globe and Mail and Estimated insured losses from the fires are between 3-7 billion U.S. dollars (Insurance Journal). According to the Washington Post, the wildfires have produced an estimated 85 million tons of carbon dioxide equivalent emissions as of 20 May 2016.

The sequence of the estimated fire perimeters can be shown through the animation below.


A subset of operations for National Weather Service (NWS) is focused on forecasting and monitoring wildfire potential and growth (i.e., Fort McMurray wildfires) during the daytime and nighttime. NWS forecasters can monitor wildfires utilizing visible imagery from satellite during the day however, monitoring wildfires during the night can become cumbersome. To assist NWS forecasters during the nighttime, polar-orbiting satellite data from the Suomi National Polar-orbiting Partnership (Suomi-NPP) satellite is considered. On-board the Suomi-NPP is the Visible Infrared Imaging Radiometer Suite (VIIRS) instrument that consists of 22 spectral channels, where one of those channels is the Day-Night Band (DNB).

The DNB provides the capability of observing night-time light emissions (e.g. wildfires) and atmospheric features across the globe and monitors the global distribution of clouds (Miller et al 2014). DNB has the ability to detect broad ranges of light intensities ranging from full sunlight during the daytime to faint atmospheric glow on moonless nights (8 orders of magnitude in radiance space). The broad range leads to difficulties in displaying an images without losing detail at either end of the radiance scale. Near-Constant Contrast (NCC) was developed to mitigate the enhancement issues utilizing a sun/moon model to convert DNB radiance values into a reflectance-like value. In short, the NCC is a derived product of the DNB.

Currently, NCC is available for NWS forecasters in the Advanced Weather Interactive Processing System-II (AWIPS-II); a weather forecasting display and analysis package. To access NCC data in AWIPS-II refer to Figure 1 below.


Figure 1: Screenshot of the AWIPS-II interface, where users can access NCC data under the ‘Satellite’ tab.

For the interested reader, a link to the Quick Guide for Imagery Enhancement involving NCC in AWIPS-II can be found below.

Furthermore, to highlight the capabilities of the NCC we will take a closer look at the Fort McMurray wildfires. Displayed screenshots of NCC imagery were taken before the fire and during the fire shown in Figures 2-4. Emitted light sources from the active fires, city lights, gas flares, and also atmospheric features such as clouds and smoke are seen in the satellite imagery.

NCC Imagery Before the Fire


Figure 2: NCC imagery was taken a week in a half before the initiation of the Fort McMurray fires at 1013Z, 18 April 2016. Note the city lights of Fort McMurray, the gas flares from Tar Island, and the clouds in the vicinity.


NCC Imagery of Fort McMurray Wildfire – 17 May at 0930 UTC


Figure 3: NCC imagery taken during the Fort McMurray fires at 0930Z, 17 May 2016. One can see the emitted light from the fires and fire perimeter line that is forming.


NCC Imagery of Fort McMurray Wildfire – 18 May at 0915 UTC


Figure 4: NCC imagery taken during the Fort McMurray fires at 0915Z, 18 May 2016. Note the active fires along the fire perimeter line and the large amount of smoke produced.


Lastly, a comparison was conducted between the estimated fire perimeter and NCC imagery for 17 May 2016 (Figure 5). Both images show relatively similar shape and size of the fire perimeter line. It is important to note the time stamps for each image in the comparison is offset by a few hours, however, the polygonal shape of the fire perimeter is still apparent.

Picture9Figure 5: The comparison between the estimated fire perimeter and NCC imagery for 17 May 2016. A polygonal shape of the active fire line perimeter is evident in both images.

Posted in Miscellaneous | Leave a comment

NUCAPS, Part One: Introduction

By Jorel Torres

The National Weather Service (NWS) has over 120 WFO (Weather Forecast Office) locations across the CONtinental United States (CONUS) where only a certain percentage of these offices produce and display RAwinsonde OBservations (RAOB). RAOB’s are important real-time observations for NWS forecasters where RAOB’s display an atmospheric stability profile, producing atmospheric measurements from the surface to the upper levels of the troposphere. RAOB’s not only assess the stability of the atmosphere, but can show levels of mixing (moist and dry air), the convective available potential energy (CAPE) needed for thunderstorm potential, determine precipitation type and can display the vertical wind profile; key parameters that are helpful for weather forecasting. The issue that comes into play is the lack of RAOB observations around the CONUS, which can become problematic for NWS forecasters. If a WFO does not produce any RAOB observations, forecasters at that WFO might rely on observations taken from WFO’s nearby, most in which are tens or hundreds of miles away. Consequently, there is a high probability the observation profile they utilize from a nearby WFO will be different than what would be seen at their own WFO; potentially causing inaccurate interpretation of real-time observations.

To help assist RAOB observations are NUCAPS (NOAA Unique Combined Atmospheric Processing System) satellite observations which combine the CrIS (Cross-Track Infrared Sounder) and ATMS (Advanced Technology Microwave Sounder) instruments on-board the Suomi-NPP satellite producing vertical temperature and moisture profiles of the atmosphere. Wherein NUCAPS and RAOB observations can be compared and displayed operationally for NWS forecasters in the Advanced Weather Interactive Processing System (AWIPS-II). Furthermore, NUCAPS can be seen as a complement to RAOB observations. RAOB observations occur every day at 00Z and 12Z only, while NUCAPS produces observations in between those hours, from 00Z-12Z and 12Z-00Z. The combination of NUCAPS and RAOB observations can further highlight the diurnal change in the vertical profile of the atmosphere since the atmosphere is always changing, and could benefit forecasters in severe weather nowcasting and the storm warning process. Additionally, NUCAPS has more observations to choose from compared to RAOB’s. Each NUCAPS observation (i.e., sounding) is approximately 50 kilometers (~30 miles) apart and are beneficial to WFO’s that do not produce RAOB observations. The differentiation between the number of RAOB observations to NUCAPS observations over the CONUS are shown in Figures 1A and 1B below.


Figure 1A: The display of NUCAPS observations across the CONUS shown in the AWIPS-II interface on 01 April 2016, @ 0846Z. Each filled circle is a NUCAPS sounding and the color dictates if the data is of good (green), ok (yellow), or bad (red) quality, respectively. It is important to note how many more NUCAPS observations there are in comparison to RAOB observations.


Figure 1B: The filled in blue circles indicate the locations where RAOB observations take place across the CONUS every day at 00Z and 12Z respectively. Notice how far away RAOB observations are from each other in comparison to NUCAPS observations.

NUCAPS expresses benefits for forecasters at WFO’s, however NUCAPS also has a few caveats. NUCAPS has trouble producing quality data when clouds are present in the atmosphere. Due to this limitation, a percentage of observations need to be modified by the forecaster, especially in the lower levels of the atmosphere where most clouds are present. Currently, this modification can be solved manually through the AWIPS-II interface, although it is cumbersome and can take too much time out of the forecaster’s daily operations. However, there are research studies that are ongoing that could help alleviate the manual modification and are working toward developing an automative process. One research study to note, is what is occurring at the Cooperative Institute for Research in the Atmosphere (CIRA) located in Fort Collins, CO, this spring and summer. CIRA is having a field campaign launching RAOB’s and comparing them to NUCAPS observations along the Colorado Front Range. The comparison of observations will be one step in the right direction in assessing what modifications are needed to produce better NUCAPS retrievals from satellite, which in turn, will increase forecaster’s confidence in its utility for weather forecasting. The field campaign starts in early May and will continue throughout the spring and summer of this year, 2016. For interested readers, look forward to future blog updates regarding the NUCAPS field experiment.

Posted in Miscellaneous | Leave a comment