A potent shortwave trough ejecting east into the central/southern plains resulted in strong surface winds and widespread blowing dust across portions of Texas and Oklahoma on 09 June 2020. GOES-East upper level water vapor imagery showed the progression of the shortwave and wrapping of dry/descending air around it’s southern and eastern peripheries (Fig 1).
NWS Norman, OK (OUN) monitored the development of blowing dust using satellite imagery and surface obs starting early in the morning on the 9th. From the OUN 1419 UTC AFD update: “Updated the forecast through tonight to add blowing dust to the weather component and lowered temperatures for this afternoon” and Aviation update: “Updated most TAFs to include blowing dust that will result in restricted visibilities and even some cigs. MVFR conditions appear likely especially for brief period across western Oklahoma.”
There are various satellite tools available to forecasters in AWIPS to observe the evolution of dust. During this event, OUN forecasters mentioned their use of IR imagery prior to sunrise and 0.64 um visible imagery after sunrise to track the evolution of the dust plume into western Oklahoma. They also noted value in using the Dust RGB, Ash RGB, and Geocolor products during the event.
During the previous late evening and early morning prior to sunrise, 10.3 um IR imagery captured the progression of the cold front south and east across the southern High Plains (Fig 2). Surface observations along the way reported the dramatic wind shift and increase to northwesterly, along with visibility reductions (due to blowing dust). The GOES imagery filled the spatiotemporal gap between those surface observations.
While IR imagery alone provides a good view of the cold air surge, it doesn’t differentiate areas of lofted dust. As has been shown numerous times on this blog, the 10.3 – 12.3 um split window difference (SWD) is efficient at detecting areas of lofted dust and areas of thick blowing dust (see here). This is primarily due to the sensitivity of the 10.3 um channel to absorption by lofted dust particles. The SWD (with IRW overlay for cold brightness temperatures [clouds]) is effective at highlighting lofted dust during the night, in addition to day. In this case, the lofted dust shows up well throughout the nighttime period as near-to-below zero values, or moving plumes of relatively dark gray (Fig 3). RGBs that include the SWD (Dust, Ash) are also effective at highlighting lofted dust at night, and are shown later in this post.
One-minute imagery was available over the region starting during the morning of the 9th from both GOES-East (Fig 4) and GOES-West (Fig 5) satellites. Comparing 0.64 um visible imagery from both perspectives, it is obvious that the GOES-West perspective provides better detection of blowing dust during this time period, due to forward scattering of the airborne particles. The high temporal coverage allows for real-time awareness regarding blowing dust location. The simple grayscale colortable is modified to focus on the lower end of the scale, better highlighting the lofted dust. This can easily be done in AWIPS by adjusting the max value downward in the “change colormap” option.
Now viewing visible imagery during the full day from each satellite, at peak sun angle, detection is degraded from both satellites given a lack of forward scattering into the satellite sensor (Figs 6 and 7). For this location, GOES-West provides superior aerosol detection in the morning, while GOES-East is best during the evening, due to forward scattering.
The Geocolor product also highlights the blowing dust, but suffers from similar deficiencies as the visible channels (Fig 8). IR bands and band combinations can then be used to improve detection.
After sunrise, the SWD-IR combo continues to prove effective in detecting the blowing dust plumes consistently throughout the whole day (Fig 9). After the initial dust plume from southeast Colorado into Oklahoma, several areas of blowing dust develop over west Texas, and later out of the OK/TX Panhandles.
Corresponding 10.3 – 11.2 um imagery captures the plumes with similar or even more pronounced negative values (11.2 um less sensitive to moisture absorption than 12.3 um), but since the clear sky areas have very small differences, contrast from plume to no plume areas is less than in the SWD imagery (Fig 10).
The Dust RGB (Fig 11) and the Ash RGB (Fig 12) contain the same ingredients (channels and channel differences), but with different thresholds. Both include the SWD, which significantly influences the dust detection. These RGBs can be used for cloud classification at the same time as dust/ash detection.
However, slight modifications to the RGB recipe(s), particularly the SWD component, can further highlight the presence of lofted dust (magenta) using the same ingredients. In this case, the only changes were to modify the ranges of the RED component to Max=0.5 and Min=-6.0, and modify the Gamma for the RED and GREEN components to 2.50. These changes take seconds to make, and can be done on the fly during an event (see tutorial here).
Wildfires also developed within the dry and windy environment during the afternoon of the 9th. The smoke and wildfires are not readily apparent in the IR detection methods used for blowing dust. However, this would be useful considering wildfires and associated smoke often develop in the same environment as blowing dust. A daytime dust/fire RGB captures the blowing dust (green), wildfire hotspots (red), and thick smoke (dark cyan emanating from hotspot) plumes (Fig 14).
Finally, during the overnight hours of the 9th, NOAA-20 and SNPP VIIRS Day Night Band Near Constant Contrast Imagery captured the blowing dust expanding across the Panhandles (Fig 15 and 16). See animation here.
Bill Line (NESDIS and CIRA) and OUN forecasters