Case Study 006"CHICAGO HEAVY
RAIN EVENT - GARP LAB"
17-19 July 1996
Written by James T. Moore, St. Louis University
You have just come on shift at the NWS Forecast Office in Chicago and it is 4:00 PM CDT (21 UTC) 17 July 1996. The tired forecaster that just briefed you has noted that conditions look dangerous in terms of a possible convective heavy rain situation. The Meteorologist in Charge (MIC) has asked that you assess the situation and prepare a mesoscale convective discussion (MCD) shortly after 7:00 PM (00 UTC 18 July 1996) tonight to update the present forecast. Using the forecast funnel approach, look at both diagnostic and prognostic data to help prepare your forecast. Remember, you cannot look beyond 00 UTC 18 July 1996 since your MCD has to issued shortly after this time. What would you forecast for the area? (Okay, so you know the outcome; at least act surprised!)
1. Construct a GOES-8, 2 km, IR satellite loop for the last 12 hours: 12 UTC 17 July to 00 UTC 18 July 1996.
Convection is occurring in the area at 00 UTC 18 July. From these imagery, is there any indication that convection may continue? What evidence did you base your answer on?
2. Examine the surface station plots for the last 12 hours, focusing on a regional scale map centered over northern IL.
Carefully look at the temperatures, dew points and winds in/near the region of interest. Especially look at the surface observations from Chicago, IL (ORD) and Quincy, IL (UIN). What does this imply about the state of the atmosphere between these two sites?
Can you detect any surface boundaries? What type of boundaries are there? Are they moving? What parameters did you use to diagnose these boundaries? Why might they be important for subsequent convection?
3. Examine the 00 UTC 18 July 1996 observed fields aloft. Focus on 850, 500 and 250 mb. You may want to use the 960718/000F000 RUC fields to get objectively analyzed (i.e., gridded) datasets. Note that RUC data are based upon a four-dimensional data assimilation of rawinsonde data, surface data, profiler data and aircraft observations (a.k.a. ACARS). The RUC is updated every three hours and run for 12 hours, however here you are looking at the initialized datasets. Especially focus on the area including Iowa, Missouri, Illinois, Indiana, and Wisconsin.
Can you diagnose a low-level jet (LLJ) around 850 mb? How strong is it? Towards what area is it directed? Will it interact with any low-level boundaries? Is there a moisture supply to fuel continued convection at 850 mb? Is there a deep layer of moisture?
At 500 mb, can you detect a short wave trough? How have the winds changed from 850 to 500 mb? Is there a cold pool of air at 500 mb that might increase the lower- tropospheric lapse rates? How moist is the air at this level?
At 250 mb, are you able to diagnose an upper-level jet (ULJ)? Is northern Illinois in a favorable quadrant of the ULJ? Why or why not? Is there any curvature to the wind field? How may this affect the divergence at this level? Where is the LLJ with respect to the ULJ?
4. Look at soundings from ILX and DVN. Which would you choose as your representative sounding (in terms of the upstream airflow into northern IL)? Why? What does the stability look like? What about the wind shear profile? How have these fields evolved over the last 12 hours?
Look at the wind profile at Slater, IA ending at 00 UTC 18 July 1996. Comment on the evolution of the wind profile at this station.
Look at fields of lifted indices, K index and precipitable water for 00 UTC 18 July. Are you impressed?
5. Now look at the Chicago radar. Set up a loop ending at 00 UTC 18 July.
Does there appear to be a fundamental change in the nature of the convective regime?
Can you note a propagation pattern developing? Where are new cells developing? Why do they favor this region? In what direction are the cells moving?
6. Now write your MCD bases on your analysis of the situation using GARP.
7. Look at the radar and satellite imagery for the period covering 0000 UTC 18 July to 1200 UTC 18 July. How did your MCD pan out?
8. Finally, look at the eta and NGM models and assess their 12-24 hour forecast (based on their 12 UTC 17 July runs) and their 12 hour forecasts (based upon their 00 UTC 18 July runs). Did they forecast the LLJ accurately? How about the low-level boundaries? Was their precipitation forecast at least in the ball park (Wrigley?)?