A mid-level trough traveling east across the northern US brought widespread strong winds and high fire danger to much of the central US on 29 Mar 2020. GOES-East water vapor imagery captures the feature, ahead of which dry and warm conditions had developed. The cold front forced south by the associated Canada/US border shortwave is also evident in the moisture imagery.

Given the expected conditions, the NWS Storm Prediction Center had issued a very large Critical Fire Weather area in their Fire Weather Outlook for the day. Widespread RH below 20% and wind gusts over 40 mph would develop within the critical fire weather area during the day.

Deep vertical mixing along with a tightening surface pressure gradient resulted in strong surface winds developing across the region. 12Z Soundings from Denver and Rapid City captured the very deep dry air ahead of the trough. SPC Sounding Climatology indicated these atmospheres very close to their record Daily Min for TPW, with 0.1″ each.

GOES-East shortwave IR imagery from the day and evening reveals the evolution of the cold front south across the plains, along with the development of wildfire hot spots (black pixels).

Focusing on South Dakota, Natural Color Fire RGB imagery from the day captures a few large wildfires and their associated smoke plumes, along with the push of cooler air.

A comparison between ABI and VIIRS Fire Temperature RGBs over the hot spots provides an example of the additional spatial detail VIIRS provides over ABI regarding ongoing burning area. A Sentinel-2 pass provided even better detail around 1751 UTC, with the 10 m true color imagery showing the burned area and smoke plumes associated with the Rapid City wildfire.

Given the strong winds, blowing dust was of concern as well. 10.3 – 12.3 um Split Window Difference imagery from the northern plains shows relatively dark gray steaks developing behind the cold front under otherwise clear skies. These are areas of blowing dust, where the SWD is relatively low, or near to below zero. Algorithms and displays with set SWD thresholds had a more difficult time capturing these plumes of blowing dust as the signal wasn’t particularly strong. Sometimes, the basic SWD with lienar grayscale colortable will be ideal for detecting blowing dust.

Combining the SWD with SWIR and IR, we have a product that can be utilized to diagnose the wildfire hot spots and blowing dust plumes together.

Another VIIRS-ABI example shows the utility of VIIRS imagery for improved assessment of blowing dust plumes, including their exact sources. A South Dakota plume is barely recognizable in the ABI Day Land Cloud Imagery, while the plume and source region are clear in the same imagery but from VIIRS. Sentinel imagery form 1 hour earlier also captured the dust plume and source with even greater detail.

A thicker region of blowing dust developed within the San Luis Valley, and east over the southeast Colorado plains. This time, we view a VIS-SWD sandwich product to diagnose the denser dust plume. Surface obs indicated widespread wind speeds over 40 knots.

Finally, a couple areas of blowing dust also developed across Utah and Nevada during the afternoon in association with a smaller shortwave embedded within the base of the broader trough.

Bill Line, NESDIS and CIRA

Strong southwesterly winds wrapping around the south/west portion of an upper low traversing the southern US plains resulted in the development of widespread blowing dust. Surface dew points within the dry southwesterly flow dropped into the single digits to below zero, while widespread wind gusts over 40 knots were also observed. Initially, blowing dust developed out of the Chihuahuan Desert of northern Mexico, and was drawn east and north across Texas throughout the day. Later, thunderstorms developing further north in east-central New Mexico sent an outflow boundary east into west Texas, along and behind which a very dense dust plume developed. This region of blowing dust, or haboob, reduced visibilities considerably along it’s path, resulting in the issuance of NWS Dust Storm Warnings. Fortunately, a GOES-East 1-min mesoscale sector was available over the region to aid in monitoring for wildfire hot spots, but also provided a unique view of the blowing dust plume. There were numerous photos and videos shared on social media captured visibilties well less than 1/4 mile within the blowing dust plume.

This was at 4:45PM taken in between Lamesa and Seminole, Texas. ⁦@TxStormChasers⁩ ⁦@rrobertswxlab⁩ ⁦@NWSMidland⁩ pic.twitter.com/oNTYdO7DrV

— Cam Wade (@gcwade) March 22, 2021

The Midland, TX NWS office noted the critical role satellite imagery and products played in operations during the event. In particular, Geocolor and DEBRA-Dust products were important in timing the outflow boundary and associated wind shift, denoted in imagery by the very thick dust plume, as it quickly approached a wildland fire west of Andrews, TX. Forecasters were able to communicate the precise timing of the outflow to local partners, resulting in fire crews moving off the southern periphery of the fire prior to the wind shift and dramatic drop in visibility, potentially saving lives and equipment. Thankfully, the wind shift took the fire south of town.

GOES-East Fire Temperature RGB Imagery within the 1-min sector highlighted the wildfire hot spot well as it developed amidst southwesterly flow ahead of the outflow boundary.

GOES-East DEBRA-Dust imagery within the 1-min sector, available on the CIRA SLIDER, highlighted the blowing dust as it developed behind the outflow boundary and approached the location of the wildfire.

Day Land Cloud RGB imagery within the 1-min sector, available in AWIPS, provided a detailed view of the very thick blowing dust plume as it developed behind the outflow boundary and advanced east through the evening (Fig 1). The animation captures the dust plume in the context of NWS Dust Storm Warnings and surface observations, which measured wind gusts over 40 knots and visiblities as low as 1/2 mile. The more transparent area of lofted dust (out of Mexico) is diagnosed ahead of the more dense dust plume that resides within the thunderstorm cold pool. The 1-min imagery allows for the exact position of the dense blowing dust and likely reduced visibility to be analyzed in real-time and between observations.

The CIRA Geocolor product was available within the 1-min sector and accessible through the CIRA Slider webpage (Fig 2). The dust plume is especially apparent in the GOES-East true color imagery closer to sunset with increased forward scattering toward the satellite sensor.

Given the dry and windy conditions, fire danger was high across the region, influencing the growth of several wildfires. One-min Natural Color Fire RGB imagery from GOES-East captured the dust plume, clouds, and wildfire hot spots (red) and smoke plumes in a single product (Fig 3). Of note is the large hot spot WNW of Midland, TX, discussed in an earlier paragraph.

The lofted dust from both Mexico and NM/TX continued east and then north around the low during the evening, and was still detectable early the next morning. A VIS-IR-Split Window Difference procedure captured the evolution of the dust during the day and evening (Fig 4). The procedure relies on the 10.3-12.3 um Split Window Difference for detecting lofted dust, and overlays it as a semitransparent shade of yellow for negative values (dust signal).

Other procedures and products capture the blowing dust day/night, including the Dust-Fire RGB (Fig 5; which also shows the wildfire hot spots), and the CIRA Debra-Dust product (Fig 6; can be accessed on CIRA Slider).

Impacted NWS offices shared various imagery products of the blowing dust, shown below.

500pm NIce satellite image of difference between an Haboob and our large scale wind/dust event. Haboob is much smaller scale but usually denser dust. #txwx #nmwx pic.twitter.com/AMTaFfLDL4

— NWS El Paso (@NWSElPaso) March 22, 2021

This satellite loop from https://t.co/LdSlQVUAo5 shows the current state of weather across W. TX. The haboob is moving into the Big Country, and a front is moving in behind it that has dropped temps into the 40s & lower 50s. #lubwx #txwx pic.twitter.com/pdoVRALTFM

— NWS Lubbock (@NWSLubbock) March 23, 2021

Very strong winds 60+ mph are causing a dust storm from SE New Mexico into the northern and western Permian Basin (Yellow on satellite). Visibilities will drop to as low as 1/4 mile or less. Please avoid travel if possible. #nmwx #txwx pic.twitter.com/hLpe2IanXI

— NWS Midland (@NWSMidland) March 22, 2021

Using our Dust RGB composite on GOES satellite, you can see those dark pink, almost red areas indicating thick blowing dust moving into Rotan and Sweetwater. Visibilities in this area have been reported as low as 150-250 yds. Avoid driving in these areas if you can! #txwx #sjtwx pic.twitter.com/phfWn3ga1F

— NWS San Angelo (@NWSSanAngelo) March 23, 2021

Procedures and products shown in these blog posts can be requested by NWS users.

Bill Line, NESDIS and CIRA

Input from NWS Midland, TX

A strong upper low digging into the southwest US forced the development of strong southwesterly winds and very dry conditions at the surface out of Mexico into New Mexico and western Texas. Winds gusted over 60 knots across the region, and dew point temperatures fell to below zero degrees F. The gusty southwesterly winds lofted a significant amount of dust out of northern Chihuahua, Mexico which quickly advanced to the northeast into New Mexico and Texas. Additionally, dust was observed emanating from White Sands in southeast New Mexico. The blowing dust resulted in a widespread drop in visibilities to below 1 mile, including reduction to below 1/4 mile at times. Given the conditions, Blowing Dust Advisories and Blowing Dust Warnings were issued by the National Weather Service.

A VIS-SWD Sandwich combo captures the evolution of the dense and wide dust plume quite well during the daytime (Fig 1). The procedure uses a VIS underlay, and semitransparent yellow 10.3 – 12.3 um Split Window Difference overlay for values less than zero. This procedure can easily be created in AWIPS (or delivered upon request) using already available data. Surface observations are included in the animation, showing the exceptional wind gusts, bone dry dew points, and reduced visibilities.

A similar animation uses products derived outside of AWIPS to capture the dust evolution. The CIRA Geocolor product provides a colored image of the scene, while the semi-transparent DEBRA-Dust overlay captures the dust as yellow through a dedicated dust detection algorithm (Fig 2). Both products are available on CIRA Slider, as well as via delivery of additional data to AWIPS.

The Dust continued to advance north and east during the overnight hours, wrapping around the eastern side of the upper low. An animation similar to Fig 1, but transitioning to IR at night, continues to capture the dust plume through the evening hours as yellow (Fig 3). By the next morning at the end of the animation, a faint dust signature is still apparent over the low clouds in Kansas.

The DEBRA-Dust product similarly continues to capture the blowing dust signature through the evening and into the next day (Fig 4).

RGBs utilizing the SWD are valuable tools for tracking lofted dust day/night, in addition to cloud classification and other applications. Additionally, weak dust signals can be captured in carefully crafted RGBs that are missed in products that include strict thresholds. An experimental Dust-Fire RGB highlights dust (day/night) as a relatively bright cyan to bright green (thick dust plume), as well as wildfire hot spots as bright red, high/opaque clouds as darker green, cirrus as dark blue or black, and low clouds as gray (Fig 5).

VIIRS I-band imagery during the early afternoon provided a very high resolution (375 m) view of the dust plume. The higher resolution imagery provides more insight into the blowing dust, allowing forecasters to more easily and more precisely identify the onset of blowing dust, dust plume edges, relatively dense portions of a dust plume, and location of dust origination. Compared in Figure 6 are the 375 m Day Land Cloud RGB from VIIRS (left), and 1 km GOES-East ABI Day Land Cloud RGB (right; at nadir, so lower resolution at this location so far from nadir).

Bill Line, NESDIS and CIRA

Forecasters at NWS Amarillo, TX utilized GOES-East shortwave IR satellite imagery to alert partners of a newly developed wildfire in the Oklahoma Panhandle during the afternoon of 04 March 2021. Mesoscale 1-min imagery was available over the region, and was utilized by forecasters to monitor for wildfire hot spots. In fact, due to periodic cloud cover across the scene, the mesoscale 1-min imagery helped in detecting the hotspot (with confidence) roughly 9-minutes ahead of what was possible in the CONUS 5-min imagery. Given the periodic cloud cover, the 1-min imagery helped to capture the “in-between” moments that the 5-min imagery missed. The utilization of the hot spot notification tool with the GOES-East 1-min mesoscale imagery allowed forecasters to relay information on the location of the wildfire and current conditions in a quick and easy manner.

A timeline comparing the 1-min mesoscale imagery (top row) with 5-min CONUS imagery (bottom row) is shown in Figure 1. Increase the Zoom in your browser for better view of the slide show. Alternatively, larger versions of the individual components of the slide show can be accessed: Period 1, Period 2, Period 3, Period 4, Period 5, Period 6. The wildfire hotspot appears as relatively dark pixels, just northwest of Guymon, OK.

An animation shares similar information in Figure 2.

Individual animations are shown for the 1-min meso imagery (Fig 3) and 5-min CONUS imagery (Fig 4). The hotspot is easily diagnosed in the 1-min imagery, despite the cloud cover, as an apparent flickering of dark pixels. In the 5-min imagery, however, there are 2 images that provide a glimpse of the hotspot, with only one showing an obvious signal.

This is a great operational example showing the value of 1-min satellite imagery (compared to 5-min) for wildfire hot spot monitoring. The availability of 1-min imagery allowed for roughly 9-minutes of lead time (over 5-min imagery) for the confident detection of a wildfire hot spot and notification of its presence to core partners.

Kaitlin Rutt, NWS AMA; Bill Line, NESDIS and CIRA