Experimental SNPP NUCAPS in BUF CWA

The Experimental SNPP NUCAPS soundings became available at 1814z with 17z validation. Overlaid on GOES-16 Red Visibility, I selected a QC approved point just off Lake Ontario near Oswego, NY.

Of note, there is a risk of cirrus contamination, but compared to the cloud free point over the eastern end of the Lake, the two profiles were similar. Of note, the latitude and longitude were 43.27/-76.26.

First, lets look at the Experimental NUCAPS sounding for this location.

What is immediately noticeable is the modestly dry mid and upper levels. with an indication of what I immediately thought was the stratus deck at approximately 1000 kft (300m). However, when looking at the 17z sounding, KFZY (Ozwego), it reported the lowest ceiling at 3300 kft, Broken conditions. This is likely due to resolution of the NUCAPS soundings, but I was happy to see it pick up on the cloud deck. Surface based CAPE was computed to be 1634 J/Kg, Low level lapse rate was 6.8 C/Km.

Next, lets compare to the 17z (initialized) HRRR sounding for the same latitude and longitude.

From the start, the noticeable difference is the moisture profile. The HRRR is more moist through the mid levels, 850 to 500mb. The temperature profile is similar, save for a weak inversion on the Experimental NUCAPS above 700mb. The Precipitable water values are very different, off by about 0.60 inches between, 1.92in from the HRRR vs 1.34in from the Experimental NUCAPS.

Now, lets compare both the Experimental NUCAPS and the HRRR to the RAP initialized at 17z.

The RAP is very similar to the HRRR profile, with identical Precipitable Water values, 1.92inches, of course this is due to the more moist mid-levels 850 to 500mb.

Quickly, lets look at the operational NUCAPS sounding, very similar to the Experimental, except it lacks the moist layer attempt to represent the stratus deck.

Final Thoughts...

I'm hesitant to trust today's NUCAPS soundings, outright, due to apparent misrepresentation of the mid-level moisture. Compared to increasing Mid-level WV across the region through 17z and more importantly the sky observations at KFZY, I have a difficult time believing this is a good representation, and this is before comparing to the HRRR and RAP soundings. Also of note, this was not a one NUCAPS sounding point issue, this was recreated over several points along the Lakeshore and northern New York.

In the end, I need more examples and experience with the profiles before completely writing off the Experimental NUCAPS, but given it's performance today, I could see forecasters losing trust in the data, especially given the time it takes to interrogate it's performance.

My other concern, for convective interrogation, either pre-initiation environments or post-initiation/mesoscale environment investigation, I worry about the utility. The timing was good, ~18z for the sounding arrival in AWIPS for most convective situations, but the static nature of NUCAPS compared to the numerous convective allowing models, that run every hour, hamper it's ability to be used during a fast paced forecasting environment. I wonder if the NUCAPS soundings would be better served as ingest into mesoscale guidance? However, in this example, it may hinder those forecasts given it's stark difference. Of course, that hinges on how accurate these mesoscale forecasts are, today.

Dale Doback