Volcanic Ash Detection and Height
Background
Given the importance of monitoring volcanic ash for aviation interests, health interests, and climate, the ABI-VAA serves a critical role in the GOES-R ABI processing system. Information pertaining to volcanic ash is needed on a very timely basis. As such, latency was a large concern in the development of the ABI-VAA. Given advances made in fast radiative transfer modeling, a state-of-the-art algorithm can be implemented without risking latency issues. The ash cloud height/mass loading retrieval utilizes the same general retrieval procedure as the ABI cloud top height algorithm. Some of the details within the retrieval procedure were modified to accommodate volcanic ash clouds, which, spectrally, behave quite a bit different than meteorological clouds. Given any type of cloud that produces a discernable signal in the infrared, the height/mass loading retrieval will produce an answer. Thus, the application of the retrieval needs to be restricted to pixels that potentially contain volcanic ash clouds. To ensure that this is the case, an ash detection algorithm is applied to all pixels prior to performing the retrieval. The ash detection simply determines the likelihood that volcanic ash is present. Volcanic ash detection is a very specialized application, so one cannot expect the cloud mask to provide this information. It is important to note that the ash detection algorithm often detects non-volcanic dust. The F&PS product statistics qualifier, "over volcanic ash cases," allows the detection algorithm to have false alarms, like non-volcanic dust.
Product Description
The ABI-VAA is responsible for producing an ash cloud height and ash cloud mass loading (mass per unit area) for all ABI pixels that potentially contain volcanic ash. A necessary intermediate product, which describes the confidence of volcanic ash being present for each pixel, is transferred into one of the quality flags.
The ABI volcanic ash products are intended to locate volcanic ash clouds and to initialize and validate ash dispersion models.
Improvements and Benefits
The Volcanic Ash product algorithm utilizes five GOES-R ABI infrared channels to automatically determine the height and mass loading properties of any pixel found to contain volcanic ash. Forecasters will be able to use the Volcanic Ash product to identify areas where volcanic ash is present and potentially hazardous, and ultimately, issue more accurate aviation, air quality, and public health warnings. It is also expected that the Volcanic Ash product will be useful for initializing dispersion models and volcanic ash trajectory prediction models. The more accurate mass loading detection may also aid in forecasting short-term climate changes due to volcanic eruptions.
GOES-R Volcanic Ash using proxy data from the SEVIRI Instrument over Iceland
How does it work? - Algorithm
The GOES-R Volcanic Ash product are generated from infrared radiances, which are day/night independent. ABI channels centered at 7.3, 8.5, 11, 12, and 13.3 μm are used in the algorithms. The 8.5, 11, and 12 μm channels provide information on cloud particle size and composition, the 13.3 μm channel detects ash cloud height, and the 7.3 μm channel detects SO2 clouds. These algorithms are unique because they account for background conditions such as surface temperature, surface emissivity, atmospheric temperature, and water vapor on a pixel-by-pixel basis. Consideration of background conditions results in greater sensitivity to thin ash and consistent algorithm performance from the tropics to the high latitudes.
See the GOES-R ATBD page for all ATBDs.
How are the results compared to existing data? - Calibration and Validation
Routine validation of the volcanic ash products is challenging since volcanic ash clouds are rarely measured by active ground-based sensors or even by active spaceborne sensors such as the Cloud-Aerosol LIDAR with Orthogonal Polarization (CALIOP). Targeted in situ measurements do not exist since it is considered highly dangerous to fly manned aircraft into volcanic ash clouds.
The general validation plan is to supplement the infrequent spaceborne LIDAR observations of volcanic ash clouds with comparisons to ash products derived from instruments that are more sensitive to volcanic ash than the ABI. Vicarious validation techniques will be used that apply the volcanic ash retrieval algorithm to other types of clouds that are commonly observed. Manual analysis will also be used to some extent.
A combination of infrared and LIDAR measurements is used to reduce the number of false positives by determining which pixels have a high probability of containing ash. Ash cloud height is directly validated using LIDAR measurements of ash clouds. Ash cloud validation is supplemented with LIDAR measurements of dust clouds which are spectrally similar, in the infrared, to volcanic ash. Ash mass loading is validated with a combination of LIDAR and infrared measures that compute a best estimate of mass loading. To the extent they are available, aircraft measurement will be used to validate the GOES-R ash mass loading product.
A more technical validation presentation, (PDF, 14.73 MB) is also available.