Using ATMS data to observe lake ice coverage over the Great Lakes

February 16th, 2021 |

ATMS Ice Concentration over the Great Lakes, from overlapping ascending Suomi-NPP passes on 30 January and 15 February 2021 (Click to enlarge)

CIMSS produces Advanced Technology Microwave Sounder (ATMS) Lake Ice concentration images in a format that can be inserted into AWIPS.  These images are created from data downloaded (available at this ftp site;  imagery is also available here) at the DB antennae at CIMSS, and processed with MIRS algorithms (all MIRS products are available at this NOAA Website) that are incorporated into CSPP.   This document (from ESA) includes a figure showing how emissivities of ice and water differ, allowing for discrimination between ice and open water.  The differences are especially large at lower frequencies.

The animation below from NOAA/NESDIS of ice concentration over the USA (including the Great Lakes) (source) shows obvious increases and decreases in ice concentration;  given the general very cold conditions over the Great Lakes during this time (especially over Lakes Michigan and Superior), the reduction in ice cover on 14-15 February is inconsistent with the cold weather.

Ice coverage increases from 10-13 February and then decreases.  This change in ice coverage matches view angle changes from the ATMS instrument on Suomi NPP, and those view angle changes affect the spatial resolution of the measurements.  There was a near-nadir afternoon pass on 10 February, and diagnosed ice in Lake Michigan was at a minimum;  ascending pass views of Lake Michigan on 12 and 13 February are near the limb and diagnosed ice over Lake Michigan reached a maximum;  the view was near nadir again on 15 February when, again, diagnosed lake ice was at a minimum.  (Suomi NPP Orbit paths are available here).  It is important when using Ice Coverage data to know the view angle from the satellite!

NPP MIRS Ice Coverage, 4 – 15 February 2021, from Ascending (afternoon) Passes (Click to enlarge)

MIRS algorithms to compute Ice Concentration use information from ATMS channels 1, 2, 3, 16 and 17.  These 5 channels have footprints ranging from 15 to 75 km (at satellite nadir), as shown in the image below, from this paper. Note especially how the footprints increase in size at the limb:  channel 17’s footprint ranges from 15×15 km at nadir to 68×30 km at the limb!   A challenge in using ATMS is that the microwave footprint can easily observe both land and water, in which case the microwave data will not give values representative of the lake coverage.

Scanning geometry for 22 ATMS channels. The figure includes footprint sizes at nadir and at the limb (Click to enlarge)

The figure below shows circles with diameters of 15, 50 and 75 km;  the smaller circle is the approximate nadir footprint of channel 16 and 17 at ATMS;  the larger circle is the approximate nadir footprint of channels 1 and 2.  Lake Ice resolution from MIRS might be considered to be of the order of 50 km.

Circles with diameters of 15, 50 and 75 km in Lakes Michigan, Huron and (inset) Erie (click to enlarge)

NOAA-20 and Suomi-NPP, the two satellites that carry ATMS as part of their payloads, both have 16-day repeats.  That is:  the satellite traces out the same path every 16 days;  in addition, paths are very similar every 5 or 6 days.  See, for example, this toggle of (Suomi NPP) NUCAPS soundings points over the South Pacific ocean, on 25 July and 10 August 2019, 16 days apart.  The same orbit is traced out on these two days.  That is why the ATMS ice concentration plots at the top of this post are from 30 January and 15 February:  16 days apart. The two orbit mappings at the links are identical. The 15 February image of orbits is shown below.

Interpretation of the Ice Concentration imagery at the top of this blog post requires knowledge about the path of Suomi-NPP shown below.  Lake Michigan and western Lake Superior are close to nadir, and there should be some ATMS footprints entirely within those lakes.  Lakes Huron, Erie and Ontario are far enough away that a user might not trust 100% the data being presented.  The ice coverage change between the two days might be useful:  there is a general increase in concentration over coastal Lakes Michigan and Superior.

Predicted Suomi-NPP Orbits for 15 February 2021 (Click to enlarge)

A morning descending pass of Suomi-NPP moved over eastern Lake Ontario, giving the best resolution over that small Great Lake.  The 0659 UTC image from 16 February is shown below.  Notice the difference in Lakes Superior and Michigan between this image (for which Lakes Superior and Michigan are near the limb) and the image at top (for which Lakes Superior and Michigan are near nadir).

Suomi NPP ATMS Estimates of Lake Ice, 0659 UTC on 16 February 2021 (Click to enlarge)

The 15 February 2021 analysis (from this page) from NOAA’s Great Lakes Environmental Research Lab (GLERL) is shown below.  Consider the ATMS imagery as an approximation to the observed field. Care in interpretation of ATMS data is a necessity because of errors that occur when pixels are not entirely over water.  That is a frequent occurrence when the satellite is scanning along the limb.

Ice concentration over the Great Lakes, from GLERL, 15 February 2021 (Click to enlarge)


The toggle below (from this site) highlights resolution differences between ATMS Channel 1 (23.8 GHz), with 75-km resolution at nadir, and ATMS Channel 17 ( 165.5 GHz), with 15-km resolution at nadir.  Note also the differences in the signals between western Lake Erie (ice covered) and eastern Lake Erie (more open water).

ATMS imagery (Channels 1 and 17) derived from Suomi NPP at ~1800 UTC on 17 February 2021 (NPP overflew Buffalo NY on this day) (Click to enlarge)

Geostationary Lightning Mapper (GLM) data displayed with Geo2Grid

January 30th, 2021 |

GOES-16 ABI CONUS sector band 13 (Clean Window, 10.3 µm) infrared imagery, 1901-2001 UTC on 30 January 2021

Geo2Grid is a scripting tool that accesses various Python packages to display Geostationary Satellite data, described on this blog before here, here and here (Polar2Grid is a similar package for Low Earth Orbit satellite data).  The animation above shows GOES-16 Band-13 (Clean Window, 10.3 µm) infrared data for an hour over Oklahoma/Kansas/Missouri/Arkansas during a time when tornadoes occurred (imagery was produced using Geo2Grid and GOES-16 level-1b radiance files).  (SPC Storm Reports).

Gridded GLM data are available at this website;  both CONUS and Full Disk domains are available, CONUS data are a simple subset of the Full Disk imagery.  These netCDF files (with ‘GLMC’ in the filename) are available each minute, and contain a variety of gridded GLM products, some of which as distributed to National Weather Service forecast offices. By using the ‘glm_l2’ reader in Geo2Grid, data can be plotted, and subsequently overlain on top of the ABI imagery, as shown below.

GOES-16 ABI CONUS sector band 13 (Clean Window, 10.3 µm) infrared imagery, 1901-2001 UTC on 30 January 2021, overlain with GLM Total Optical Energy at 1-minute time steps (Click to animate)

NUCAPS fields across an upper tropospheric front

January 20th, 2021 |

GOES-16 ABI Airmass RGB, Band 10 and Band 8 (7.34 µm and 6.19 µm, respectively), and GOES-16 Airmass RGB overlain with NUCAPS sounding availability plots, 0801 UTC oni 20 January 2021 (click to enlarge)

The AirMass RGB from GOES-16 at 0800 UTC on 20 January 2021 showed a distinct color change across central Missouri, from red to green.  The enhanced red coloring suggests a large difference in water vapor brightness temperatures.  The toggle above (including an image with NUCAPS* sounding points), shows structures in the water vapor imagery consistent with an upper tropospheric front.

Water Vapor and Airmass RGB imagery fields are useful because they be compared to model fields of the tropopause, and similarities in model fields and satellite imagery lend credence to the idea that the model initialization is accurate.  Compare the Airmass RGB and the Rapid Refresh mapping of the pressure on the 1.5 PVU surface below.  There is good spatial correlation between model and satellite fields.

GOES-16 Airmass RGB and Rapid Refresh model field of Pressure on the 1.5 PVU surface, 0800 UTC 20 January 2021 (Click to enlarge)

How do vertical profiles from NUCAPS vary across the tropopause fold?  The animation below shows six different profile in Missouri and Arkansas, spanning the reddish region of the airmass RGB.

GOES-16 Airmass RGB image with selected NUCAPS profiles, as indicated. (Click to enlarge)

A more efficient way to view information from NUCAPS is to view gridded fields.  Polar2Grid is used to transform the vertical profile to horizontal fields at the individual NUCAPS pressure levels (and then vertical interpolation moves those fields to standard levels).  The animations below show gridded values that are all in agreement with the presence of a tropopause fold where the Airmass RGB and model fields suggest.  Gridded temperature and moisture can be combined in many ways.  Gridded Ozone is also available in AWIPS (some of these fields were created using the Product Browser).

Ozone from NUCAPS, below, does show an enhancement, as expected, in the region where the tropopause fold is suggested by the airmass RGB.

NUCAPS-derived ozone anomalies, ca. 0800 UTC on 20 January 2021 (Click to enlarge)

The gridded NUCAPS tropopause level, shown below, can also be inferred from the individual profiles shown above.

Gridded NUCAPS Tropopause level, ca. 0800 UTC on 20 January 2021 (click to enlarge)

Note how the lapse rates show relatively less stable air (in the mid-troposphere) in the region of the tropopause fold.

Gridded 500-700 mb Lapse rates, ca. 0800 UTC on 20 January 2021 (click to enlarge)

Mixing ratio shows dry mid- and upper-tropospheric air, in the region of the tropopause fold, as might be expected from the GOES-16 water vapor imagery.

Gridded NUCAPS esimates of 300-700 mb mixing ratio, ca. 0800 UTC on 20 January 2021 (Click to enlarge)

In general, NUCAPS data can be used to augment other satellite and model data to better understand the thermodynamic structure of the atmosphere.  For more information on NUCAPS profiles, refer to this training video.

*The careful reader will note that the timestamp of the NUCAPS Sounding Availability plot, 0753 UTC, is different from the GOES-16 imagery.  Why?  The NUCAPS Sounding Availability plot is timestamped (approximately) when NOAA-20 initially overflies North American airspace.  NOAA-20 was flying over Missouri shortly after 0800 UTC, as shown in this plot (from this website).  Gridded NUCAPS fields are timestamped when NOAA-20 is overhead.

CIMSS GeoSphere website is now active

January 13th, 2021 |

CSPP GeoSphere page showing True-Color (and Nighttime Microphysics) Imagery at 1730 UTC on 13 January 2021 (Click to enlarge)

The CIMSS CSPP-Powered GeoSphere site is now available.  This site allows quick access to GOES-16 imagery via Cloud-friendly, flexible (and configurable) software.  Data are accessed from the GOES Re-Broadcast (GRB) downloaded via antenna at CIMSS/SSEC.  Significant effort was made to reduce latency.  For example,  GOES-16 Mesoscale Sectors load within about a minute of their being downloaded, CONUS sectors load within about 8 minutes, and Full-Disk imagery loads within 20 minutes.  In addition to the individual bands (1-16), a sharpened True-Color (Daytime) and Nighttime Microphysics (Nighttime) product is available.  The data are tiled and only tiles that are needed for the present view are loaded.  In addition to still images, animations can be displayed, and a user can choose how many images are in the loop (up to 30), and what the time-step (i.e., ‘Pattern Stride’) is.  Users can also share urls so that others can view the same animation.  For example, the image above was created with this url.  Individual frames can be downloaded via a left-click on the image.  There is at present no method to download animations directly.

More information on this site is available in this recorded presentation created from this PowerPoint.