Evolving the National Weather Service. 3. What the last fifteen years have to tell us with respect to the technology piece.

…continuing from the previous LOTRW post…

Now we find ourselves in 2015. It’s Groundhog Day. Déjà vu. The WSR88D’s, once such an innovation, have themselves seen service for almost 30 years. The rotating radar mounts are wearing out. Technology is again antiquated. The NWS is engaged in a NEXRAD service-life extension program. While important, the experience of the past shows it at best only delays the inevitable: the need for a complete rework of the system. And it’s delaying a safer American future. Today’s phased-array radar technology uses digital circuitry to scan the skies vs. relying solely on physically-rotating-and-steering radar mounts. If implemented nationwide, it would offer faster scans and much-more-frequent updates on the threats posed moment-by-moment by rapidly developing storms.

ASOS, the automated surface observing system that had previously seemed so new, now needs, and is getting, a similar facelift. After many upgrades, AWIPS is being replaced by AWIPS II. Modelers at the NWS National Centers for Environmental Prediction once more have their hands full introducing step-function advances in numerical weather prediction. The introduction of these and other changes over the next few years will require that extensive technological advance occur side-by-side with day-to-day provision of NWS services, while at the same time reconfiguring the job of the NWS bench-level forecaster.

Thus 2015’s “evolving the NWS” looks very similar to yet another MAR, only called by another name.

Importantly, there’s much here to celebrate. Major advances in public weather services are in the offing. And important new public-private collaborations to better serve the public are waiting in the wings as well. But intermittent, crash modernizations and catch-up implementation prompt searches for be a better way. Had the past fifteen years gone differently, might not some if not most of this capability could already be in place? That had certainly been the aspiration throughout. Everyone responsible wanted to see a more orderly, continuous process of innovation and upgrades. Why hadn’t that come to pass?

Significantly, the root causes (and the fixes) lie both outside and within the agency.

Let’s start with a closer look under the hood of the last century’s MAR. Why did it succeed? Here’s one reason. By the late 1970’s, NOAA leadership recognized that it needed to develop new forecaster workstations but lacked a coherent, robust set of requirements to design against. The design of a suitable forecast workstation posed substantive research questions – not just about meteorology, but also about forecaster-human-factors, and about the technology and social science of risk communication to the public. Once that design could be determined, then taking it from research into operations (R2O) would still require additional steps. To accomplish the needed R&D, they established the Program for Regional Observing and Forecasting Services (PROFS) within the research arm of NOAA. Happily, NOAA/PROFS scientists and engineers didn’t rely on brute force to accomplish the work. They used finesse, embodied in three remarkable innovations[1].

First, they designed and built a so-called Exploratory Development Facility or EDF, which allowed different workstation configurations to be programmed up using mere software. The first such workstation took two years to build because of the insertion of this additional step, but subsequent iterations and redesigns, depending upon the nature of the changes, could be accomplished in days or weeks. In effect, PROFS engineers and meteorologists had invented rapid prototyping before it became widely known by that name in other fields of applied science and engineering in the late 1980’s.

Second, PROFS brought in groups of operational forecasters to test alternative workstation configurations and provide the feedback needed to derive a nearly optimum system. They embedded researchers with practitioners and vice versa. By blurring distinctions between the two, they not only dramatically sharpened communication and refined the final product but in the process improved the morale and sense of purpose of both groups.

Third, PROFS introduced a management innovation. Though housed administratively within the research arm of NOAA, PROFS was jointly managed by a so-called troika – the leadership of NWS, the Office of Oceanic and Atmospheric Research, and the National Environmental Satellite and Data Information Service (NESDIS). Staff from these three NOAA Line Offices were in daily communication, and in many cases collocated for the work at hand. The three associate administrators would meet in person for two-day quarterly reviews with several levels of staff working on the project.

In retrospect, the secret sauce in all this was the choice to focus initially solely on workstation design rather than dissipate that same attention over the whole of the MAR. The forecaster workstation was both a key NWS outlet to public service delivery (especially public-safety messages) and a key customer for all the upstream observations and centralized numerical weather prediction products. It was in effect impossible to achieve success with respect to the workstation design without getting all the other links in the service delivery chain properly aligned. Management attention steadily expanded to include these larger considerations.

In the end, PROFS and its EDF shaved years off the time required to develop what would become the Advanced Weather Interactive Processing System (AWIPS) system and greatly enhanced the final result.

This background provides a partial list of teachings from the MAR that might be applied moving forward. Here’s a start (you can readily come up with your additional ideas and better formulations):

• Accelerate and strengthen the technology infusion; be as disciplined about, and give as much priority to, R2O as to either research or operations per se.
• As a starting point, continue to invest in the work and capabilities of PROFS – the rapid-prototyping capability developed during the late-20^th-century MAR that not only greatly accelerated development of the forecaster workstation but ensured that it would be tailored to forecaster needs. Today, PROFS exists under a different name – the Global Systems Division of NOAA’s Earth Systems Research Laboratory. It’s functioning well but its reach is constrained. This is not for lack of ideas but by funding and staff levels. Moreover, the multiple-line office, AA-level attention and priority given the work has fallen off its (intense) levels of the 1980’s and 1990’s.
• More fundamentally, NOAA’s struggle to stay abreast of rapid advances in science and technology show the need to weave R2O capability throughout the fabric of the entire agency. Pockets of excellent R2O exist here and there across NOAA. But what’s needed is a robust R2O infrastructure connecting the successful bits and filling the gaps, and a culture of R2O guiding the whole. (In this respect it’s heartening to see NOAA leadership and Chief Scientist Rick Spinrad in particular taking a hands-on, agency-wide leadership role with respect to R2O across NOAA.)
• Lastly, NOAA should remember that it is fundamentally a science and technology agency. Here’s an important distinction: Change – especially technological change – happens to agencies such as, say, the Social Security Administration. The technology for providing benefits to the elderly may change but not the basic purpose. By contrast, NOAA and its Line Offices are constantly reinventing themselves and what they do. That dynamic is inherent in problems like building weather-readiness and managing marine resources and monitoring the Earth. NOAA effectuates change, embodies change, even as it responds to changing external circumstances, both natural and social.

That brings us to the outside piece. NOAA cannot accomplish this alone.

This means that the late-20^th-century MAR also holds lessons for members of Congress, and for other stakeholders outside NOAA and the NWS as well. The nation needs to put priority on innovation and on its application for societal benefit.

• Most immediately, it’s shortsighted to nickel-and-dime the NWS and settle for life-extension plans for outmoded systems. Any apparent cost-savings are illusory. Instead, Congress and the executive branch should invest in the new technologies and networking opportunities. In the short term this puts Americans in both public- and private sectors to work at high-tech tasks that will build 21^st-century infrastructure. In the intermediate term this creates global markets for American products and services. And in the long term it builds an America more resilient to natural hazards at home and a good neighbor when it comes to building weather-ready nations abroad.
• As a starting point, Congress and the executive branch could accelerate the adoption and implementation of phased-array radar technology for monitoring storms.
• Congress might also work with NWS and the Department of Commerce to develop and strengthen the modalities that allow the NWS to work with the private sector at all levels, ranging from aerospace corporations to local small businesses, to make co-production of meteorological knowledge, vigorous R2O, and continuous infusion of new science and technology across the entire government and corporate enterprise a reality, not just a mere aspiration.
• Although much work is needed, recognition of this need is widespread, particularly with respect to development and operation of satellite-based observing systems and harnessing of so-called big data. But another starting point that merits additional attention is the so-called National Network of Networks. NOAA controls and operates only a few percent of the surface meteorological instruments across the United States. Most are operated by state and local governments and the private sector. To coordinate these observations and build a whole more effective and efficient than the sum of the parts is as much political and legal challenge as it is technical.

The next post will focus attention on the workforce dimensions of last century’s Modernization and Associated Restructuring, which were profound.

_________

[1] Much of the credit for this vision and the early implementation goes to two remarkable individuals, whose names wouldn’t be recognized by most NOAA professionals today: Donald W. Beran, who was the first director of PROFS, and David Small, who was his senior engineer. Later it would be Alexander (Sandy) MacDonald who would draw on his unique background spanning Air Force forecasting experience and meteorological research who would take over leadership of the program and really make it go.