The moving shadow of Denali

March 12th, 2021 |

GOES-17 CIMSS True Color Imagery, 1650 UTC 12 March – 0230 UTC 13 March 2021 (Click to enlarge)

Clear skies over much of Alaska on 12 March meant that GOES-17 had a clear view of Denali.  The animation above, of GOES-17 CIMSS True Color imagery, shows the shadow of Denali (near the center of the imagery, due north of Cook Inlet) moving during the day.

A higher-resolution animation, below, using only the 0.64 µm visible imagery, from 1610 UTC on 12 March to 0240 UTC on 13 March 2021, also shows the changing of the shadow during the day.

GOES-17 Visible (0.64 µm) Imagery, 1610 UTC 12 March – 0240 UTC 13 March 2021 (Click to enlarge)

Using NUCAPS data to judge model performance before a storm

March 12th, 2021 |

NUCAPS representation of tropopause height (10:26 UTC, in millibars) and Rapid Refresh model estimate of tropopause height (1000 UTC). Click to enlarge)

A potent winter storm is on tap for eastern Colorado (and surroundings) over the weekend.  As with any system bringing heavy snow, knowing the storm path is crucial.  From a recent Boulder WFO Forecast Discussion, available here:   “Recent models runs in the GFS and ECMWF have trended more to the north while the Canadian stays to the south.”  The toggles above and below compare NUCAPS satellite-based observations of tropopause pressure (NUCAPS at 1023 UTC above, at 0843 UTC below) with 40-km Rapid Refresh model estimates (1000 UTC above, 0900 UTC below) of the tropopause pressure.  Can you use the NUCAPS depiction of the tropopause to convince yourself that the model — the Rapid Refresh in this case — has the proper initialization/evolution of the impulse that will generate the snowfall?  (Note that the colormaps for all images are the same).

NUCAPS representation of tropopause height (08:44 UTC, in millibars) and Rapid Refresh model estimate of tropopause height (0900 UTC, in millibars). Click to enlarge)

A challenge with this storm at this time is that the neither of the NOAA-20 NUCAPS fields available in AWIPS themselves sample the entirety of the tropopause fold.  One might use the 0844 UTC imagery above to conclude that the Rapid Refresh is too slow in the eastward progress of the storm.  It’s difficult to make the same conclusion from the 1026 UTC image, however, at top.  Ozone anomalies (shown below) from the NASA SPoRT gridded NUCAPS site (link) include the Suomi-NPP pass (in between the two NOAA-20 passes above) that does more completely sample the tropopause fold at one time;  however, those data are not available in the baseline AWIPS.  Perhaps the afternoon pass from NOAA-20 will offer better coverage.  (Update, below:  It did!)

Gridded NUCAPS estimates of ozone anomaly at 0840 (NOAA-20), 0931 (Suomi NPP) and 1023 (NOAA-20) UTC (click to enlarge)

Gridded Tropopause height fields and gridded ozone fields, shown below, do agree very well; use them interchangeably to identify tropopause folds.  The two fields are derived from different CrIS channels in NUCAPS.

Ozone concentrations and Tropopause pressure from NUCAPS, 0844 UTC on 12 March 2021 (Click to enlarge)


The 2000 UTC NOAA-20 overpass more completely sampled the upper tropospheric feature.  See below.  The placement of the feature in the Rapid Refresh looks approximately correct over the southwestern United States (that is, NUCAPS observations seem to overlap Rapid Refresh features);  there are some dissimilarities in features over the Dakotas:  farther north in NUCAPS, farther south in the Rapid Refresh.

NUCAPS representation of tropopause height (20:05 UTC, in millibars) and Rapid Refresh model estimate of tropopause height (2000 UTC, in millibars). Click to enlarge)