Mike, I made many points and you seemed to skip over a lot of them...
If it is vital to initialize the mesoscale models (NCEP does not seem to agree since they frequently do not run or are far behind others), then SGF (to use May 22 as an example) can send up a rawinsonde every 1 or 2 hours whenever the very rare combination of >4000j of CAPE and SRH >175 presents itself. That would better (since temperature, humidity and pressure would be added) analyze initial conditions for the model than radar alone. The more distant locations (i.e., Pittsburg, KS, Lake of the Ozarks, West Plains, FYV, etc.) under the radar's umbrella would still be measured by the lower tilts are they are now.
A single rawinsonde gives information at a (theoretical*) point, whereas the radar can give this information over a much larger area. Furthermore, a rawinsonde every hour does not help with using EnKF for initializations. WoF work has shown that it takes a certain number of radar scans for the model to "take" it, this is one of the benefits of MPAR -- we can get the required number of volumes in less time. In any event, this is admittedly the weakest argument. I merely included it as an example of other purposes for volume scans than composite reflectivity...
* I said "theoretical point" because rawinsondes get ingested into the model at a single grid point, even though they are being advected around by the environmental wind and are not guaranteed to be valid at the point they are representing in the model. This is especially true near atmospheric boundaries -- where a few kilometers can make a huge impact on what the sonde observes.
You do not need to spin the radar faster to get more low-level coverage if you only use the lower-four tilts. No additional wear and tear.
You do if you want to keep the higher tilts, as I am advocating for. I use those higher tilts quite frequently -- even in tornado environments. A lot of useful and important information can be gleaned from them if people are willing to use them.
I'm not advocating doing this in the situation you pose above. Presumably, a local NWS office is sufficiently threat-aware to know the difference between a May 22 and, say, the giant hail storm in ICT on Sept. 18, 2010.
But what about a case like 10 May 2010 where Norman was under the gun with what became an EF-4 tornado, and 5"+ hail was falling in north Norman and Moore? Don't the residents of Moore deserve to be given a heads up that 5"+ hail is coming? Furthermore, environment aside, how can you tell from looking at radar data how strong a tornado might be? I've seen plenty of rotation signatures that were stronger than the Joplin one that didn't produce tornadoes. I've also seen signatures that were weaker produce strong tornadoes. Where do we draw the line? Maybe with environments like 27 April 2011 it might be easy to know, but the environment on 22 May 2011 was hardly screaming EF-5 tornado. And on 27 April 2011, what benefit would more frequent low-level updates have been? In that kind of environment you tend to think tornado warning first, severe second. Sure we'd be able to see the parent circulation move through cities with more temporal precision, but how would that have impacted the warning services?
Second, I don't understand the seeming preoccupation with liability among government-employed meteorologists. It is very hard to get permission from a federal judge to sue the federal government and, almost certainly, a decision as to how the radar is run would fall under the federal government's "discretionary function exemption" to the Federal Tort Claims Act that says that the government cannot be successfully sued for using good-faith discretion (choosing one volume scan over another) in carrying out its operations.
First of all, I'm not a government employed meteorologist, so I don't know what that first line is intended to suggest. Secondly, I never once mentioned lawsuits as that thought had never entered my mind. What about service assessments? Congressional inquiries? Investigations by the Inspector General? Even though people might not "lose" lose their jobs in the government over making poor decisions they can certainly be pushed out? Don't believe me? Ask the former MIC in Tulsa what happened to him after the 21 April 1996 Fort Smith, AR tornado.
At the very least these "investigations" are time consuming and a drain on all involved. They take away people from their normal job and tend to be a factor on the overall budget. So, yes, I think it is valid to be thinking about these things.
If your concern is about private sector meteorologists, this scenario wouldn't give me two seconds' thought. I wouldn't see any liability at all from choosing to monitor for tornadoes more closely than a hailstorm in a situation similar to May 22, 2011.
I'll be honest. I'm getting really frustrated at the number of people talking about how great the environment on 22 May 2011 was and how it was obvious to know that a major tornado was going to happen. Sure, in hindsight we know what happened, but that environment happens a lot more frequently than people realize without major tornadoes occurring. How are we to know that this time it's the real deal?
But let's assume for a second you are correct and we had turned the radars on for more low-level sampling. What good would that have done? Other than letting us see every 2-3 minutes instead of 4-5 that a tornado was possibly going through Joplin, what could we have done differently? Sure, maybe you could provide more specific warning guidance, but chances are those in the path won't get it as the power will probably have already been cut. Furthermore, do you trust the radar enough to tell people 2 miles north of the tornado they are safe? I certainly don't -- especially at the ranges most people are covered by the radars.
More frequent low-level updates have the greatest impact in marginal settings when a forecaster is on the fence regarding whether a tornado is developing -- not when the radar signature is already quite intense. So, once again I ask, how are you going to know ahead of time what impact more frequent low-level updates will have?
Ultimately, it is my opinion, that if you want more frequent low-level updates, and you are going to argue that aviation forecasters want the same, then re-evaluate the TDWR program. These are specifically designed for that purpose.
