New Report: A National Strategy for Advancing Climate Modeling
In late 2010 and 2011, I was writing about organizing U.S. climate modeling. I combined and posted some of the WU blogs on ClimatePolicy.org as Something New in the Past Decade? Organizing U.S. Climate Modeling. I want to revisit those issues in light of the release of a National Academy of Sciences Report, A National Strategy for Advancing Climate Modeling (2012).
I am a co-author of this Academy report. In this blog, I am writing not in my role as a co-author, but from my personal perspective. This blog fits in with many of the themes I have written about in the last few years.
First, I want to explain the role of the National Academy of Sciences. The Academy is a private, not-for-profit organization created by President Abraham Lincoln at the height of the Civil War. Lincoln and others at the time realized the importance of science and technology to the United States and wanted a way to get independent advice on issues important to policy. Almost 150 years later, this importance is greater, but the role of science is an increasingly controversial political issue – especially when scientific investigation comes into conflict with how we might want to believe and to act. (see, here or edited here ) So one role of the National Academy is independent review – a role that is at the heart of the scientific method and the culture of scientific practice.
Second, how does the Academy decide what to write about? The Academy serves as adviser to the government, and so organizations within the government ask the Academy to evaluate a specific set of questions or issues surrounding a body of science-based knowledge about a particular subject. Often, as in this report, there is a forward-looking aspect of the problem, such as an outline for a strategy. One example of a past report is an analysis President George W. Bush requested soon after his inauguration, Climate Change Science: An Analysis of Some Key Questions. This report is famous, partly, because it took one-month from start to finish. It found that the science of climate change was robust. In June 2001, President Bush gave a speech noting, “Climate change, with its potential to impact every corner of the world, is an issue that must be addressed by the world.” (see speech)
The current report on modeling strategy had many sponsors, notably among them the U.S. Intelligence community, a user not a builder of climate models. An early talk to the writing panel was given by Rear Admiral David Titley (Titley talks about climate and national security). In his presentation, he highlighted numerous concerns of the U.S. military, ranging from patrolling an open and disputed Arctic Ocean to threats of rising sea level to billions of dollars in assets. Other issues of national security are related to the stability of nations, the access to resources, and the volatility of commodity markets. A basic question to the panel is on the improvement of predictive skill, to address questions such as, when will we have to rebuild the dry-dock in Newport News, Virginia, and how high will it have to be?
Next, how does a panel like this actually work? The National Research Council is the operational part of the Academy. If you look on the Policies and Procedures link, you will see its rules of operations on, say, conflict of interest. The Academy selects a chair and a panel to answer the sponsor’s questions from a broad range of experience and points of view. Practically, members of the panel are assigned as lead authors on some chapters, secondary authors on other chapters, and reviewers and deliberators on the entire document. In addition to the panel, the Academy assigns staff members to manage the integration of the document as well as to assure the document is written according to Academy protocols. The staff is attentive to moving the document away from personal points of view towards a document that represents the collective view of the panel. That’s the process.
In this report I wrote history – the first draft of Chapter 2, Lessons from Previous Reports on Climate Modeling. Also, having been a co-author on some of those earlier reports, I provided continuity. For this blog, I am going to write from the perspective of someone who has advocated the need for our community to address a set of important organizational challenges. Or given a more than 20-year history of repeated recommendations and a series of Academy reports that re-identified the same problems, as stated in Chapter 2, “A challenge, therefore, to the current committee is how to disrupt the inertia of the U.S. climate science enterprise: going forward, what do we do differently?”
Because of the disruptive consequences of global warming, the scientific study of climate change has, long ago, moved out of the domain of curious scientists driven to explain the world around them. Climate change requires more than interpretation and guidance in order to be relevant to policy. Stated differently, to be directly usable by society, there is a requirement for scientific investigation focused on specific questions or classes of problems. Addressing these problems requires the use of complex software systems, multidisciplinary scientific information, rigorous and transparent evaluation, and interpretation of the knowledge produced and its uncertainty. Therefore, addressing these problems requires the combined efforts of many individuals from several professional backgrounds. There needs to be a process of planning, coordination, and execution.
We need, therefore, to coordinate activities that are, traditionally, scientific, computational, and organizational. My experience as a manager of scientific efforts is that organizational coordination is far more difficult than the challenges of computational and scientific coordination. Standing alone, coordination of computational and scientific efforts is stunningly difficult. Therefore, the new, perhaps overarching, recommendations of this report are focused on ideas that the committee viewed as helping to advance coordination, integration, or synthesis.
One of the report’s overall recommendations is “to evolve” towards a national software infrastructure for climate modeling. I think the word “evolve” is important because the reports from a decade ago also recommended software and information system infrastructure. In fact, following those reports, there has been investment and progress, both substantial, in the development of infrastructure. This is documented in the report, with the recognition that the organizational achievements are as notable as the technical achievements. Throughout the report, there are calls to build upon these successes, to utilize the communities that have made the progress of the past decade. To quote, “The Committee recommends a community-based design and implementation process for achieving a national common software infrastructure. While this goal has risks, costs, and institutional hurdles, the Committee believes they are far outweighed by its benefits.”
Another major recommendation is the formation of a modeling summit to promote “tighter coordination and more consistent evaluation” of climate models. This, to me, is perhaps the most novel and most important recommendation. Why? Previous reports have struggled with organizational issues and have made recommendations about re-organizing government agencies or re-focusing governmental organizations. At the same time earlier reports, as well as this report, express reservations about centralization and bureaucratic structures. What this recommendation recognizes is the need for a community-based organization that needs to find its niche within the federal agency structure, the interagency organizations, and the growing community of users. It recognizes the value of increased community-based planning and, hopefully, execution. And it, once again, recognizes the progress of the past decade of community building.
The next key recommendation is to “nurture a unified weather-climate modeling effort that better exploits the synergies between weather forecasting, data assimilation, and climate modeling.” This subject, too, has been flirted with in previous reports, and it is a recommendation that is more controversial than one might imagine. These two communities, weather and climate, have come to the modeling problem from different perspectives. Their practices of science have some distinct differences. There is also in the United States an idea held by many that weather forecasting is “operational,” and that “operational” comes at the expense of “science.” This recommendation from the Academy panel is based on the facts that 1) “operational” does not have to come at the expense of “science,” and 2) rationalization or unification of the different practices of science come with the benefit of more robust science-based products.
The final overarching recommendation is about the development of a new type of professional, the climate interpreter. This recommendation follows from other Academy reports and a growing body of research into the barriers of the use of climate information by scientists and practitioners who need climate information in their research, applications, and decision making. This recommendation explicitly recognizes the importance of formalizing the interfaces between climate modeling, more broadly climate science, and the usability of climate information by society as a whole.
These new recommendations are supported by a series of recommendations, which are, again, focused on pulling together the community: the scientific efforts, the computational efforts, and the interfaces to society as a whole. These supporting recommendations focus on continuation and strengthening of important activities that are of especial importance.
I want to also point out a few things that the report is not. It is not a list of important scientific questions. Many such lists have been made, and they are often the natural product of a group of scientists thinking about strategy. It is not a recommendation that if the government reorganizes in some way or simply provides more money, then we will address all needed climate services. We have no way to reorganize the government, and we are smart enough to know the challenges of money. And, finally, this report is not a call to centralize through reorganization. As a government manager, for years I studied centralized organizations, federations, and anarchist groups. I feel that centralization in a field and environment like ours is, primarily, a process that leads to increased risk. I feel that federated, community-based responsibility is the best path to assure success. It is also the most difficult.
As a final comment: I am a co-author of the report writing a blog that is my point of view. If I were asked to interpret the report in a strategic sense for a program manager, this is where I would start. One of the lessons I have learned is that this report is now in the hands of the public. Some people will interpret the report to support their agendas, sometimes their prejudices. I looked at past reports that I have been involved with, and I have seen recommendations cherry-picked for both good and bad reasons. The message of this report is synthesis, integration, and coordination. For the report’s message to become reality, those with the power to act and to implement need to focus on synthesis. We need to go forward more as a whole than as a thousand points of expertise brought together in grand exercises of climate-science assessment.