“the war to end all wars” – British author H. G. Wells (referring to World War I).
“This war, like the next war, is a war to end war.” British prime minister Lloyd George, at war’s end.
(Continuing lessons from history as they inform the current NOAA effort to evolve the NWS…)
World War I was horrific in terms of casualties and geographic reach. The carnage led some to hope that nations and their leaders would forever after forego war as a means to settle differences or achieve aspirations. However, by the conflict’s end, the world had grown more cynical and less sanguine.
Something similar happened with the NWS Modernization and Associated Restructuring (MAR) late in the 20th century. Here’s the (greatly over-simplified) story.
One driver for the MAR – in many ways the poster child – was the need to replace the aging network of WSR57 weather surveillance radars (pictured). The designator “57” referred to the year 1957 when the network had been deployed. The technology was in reality much older still – dating back to World War II.
By the 1980’s, radar capabilities had advanced. Any replacement radars could have Doppler capability, recognized as a sine qua non for detection of thunderstorm rotation and early warning of tornadoes. All that would be required was a bit of extra signal processing electronics, representing an additional cost amounting to no more than “sales tax” on top of the cost of several million dollars for each unit. Dual polarization, which could unambiguously distinguish between hail and heavy rain – was also at hand, and would have cost no more than another $10K per unit.
But it was mere upkeep – not any futuristic vision for improved public safety – that was driving the MAR. WSR57 electronics relied on vacuum tubes. Some were so outdated that when they burned out (much like any light bulb) they had to be painstakingly dismantled and refurbished by hand; they were no longer in manufacture. Aging radar mounts were also failing. No longer able to rotate as intended, they were seizing up on their pedestals at weather stations across the nation. Radar outages could therefore be numerous and prolonged. The problem was clearly set to grow with every passing year. In fact, when NIST (yes, that NIST) did a cost-benefit study for replacement of the radars, economists were able to justify the total cost of the MAR as a whole entirely on the basis of maintenance costs saved just on the WSR57’s alone.
The NEXRAD (NEXt-generation RADar) system of WSR88D weather surveillance radars was the result – finally installed beginning in 1988, 30 years after the old system had been put into operation – with Doppler (hence the D), but lacking the additional hail-heavy-rain discrimination feature.
By the 1970’s and 1980’s, radars represented only one of many systems that had grown outdated. The NWS also needed to automate surface measurements of temperature, humidity, wind speed and direction and how the data were shared system-wide. In Weather Service Offices of old, the radar display might sit in one corner of the room. Satellite imagery might be coming in by fax in another corner. Products from the primitive computer models of the time would come in by fax as well. Warnings might be issued from the office via individual telephone calls to broadcast media and emergency managers who could then get the warnings out more broadly. The communication was analog. But by the 1970’s and 1980’s, computers and digital communication were making it possible and economic to automate all of this. In principle, forecasters should be able to sit at a single desk, with multiple displays integrating models, satellite, radar, and surface information, generating and disseminating warning text and products – all with just a few key clicks. The result would be development and installation of the Advanced Weather Information Processing System, or AWIPS. This became an NWS/MAR goal.
Like World War I (but, happily, with fewer actual casualties) the MAR proved wrenchingly difficult and painful for all NWS employees from headquarters down to the field forecasters. All the new technology had to be developed without skipping a beat putting out the hourly and daily forecast products – and keeping up this dual workload for something like a decade. At the same time, the new technology was enabling and also forcing changes in forecaster workload and role. During this period, the job of NWS deputy director morphed into the toughest job in NOAA. Then-NWS-Director Hallgren split that role into a deputy for day-to-day operations and a deputy for the transition, instantly creating the two toughest jobs in NOAA.
At the end, all NWS employees involved – at all levels – breathed a collective sigh of relief, and in the same breath, said, Never again! This may not have been the first NWS Modernization and Restructuring, but it will be the last!
(Unconsciously channeling H.G. Wells) This was to have been the MAR to end all MARs.
Instead (said these same NWS personnel at that time), from now on, we’ll make such draconian, disruptive, step-function modernizations unnecessary. We’ll continuously infuse new technology into services as soon as it becomes available.
But a different reality quickly set in. Before long, day-to-day operations had once again become the major focus. Infusion of new science and technology has since been continuous but slow, failing to keep pace with the boom in science and in information technology over the past 15-20 years.
Why did the outcome everyone wanted to avoid – the need for yet another technological and structural upheaval, versus a more gentle, continuous process of innovation – happen anyway?In the next post, some lessons from this narrative…
________________
A postscript. The National Weather Service struggles of the past 15-20 years don’t reflect on any organizational lack of competence or effort, or leadership lack of vision. Nor are they in any sense unique. Instead they mirror nearly-universal travails of all large-technology-based enterprises as they try to stay current. Take, for example, the U.S. military. You can find an excellent summary of this parallel history and the transcendent stakes for America’s standing in the world in this article – Who’s Afraid of America? in the June 13th print edition of The Economist.)
