- NWS high-resolution forecast of Irene’s winds for 1 a.m., Sunday, Aug. 7.
I started to write this last week in response to a story that appeared in the Aug. 22 Washington Post. The story discussed potential funding cuts for polar-orbiting weather satellites, meaning those that orbit the earth over the north and south poles and thus as the earth rotates observe the entire planet, and here's what I was correctly quoted as saying:
“I think the loss of any data will have a negative impact,” Ryan said. “You can pick and choose big, significant events and worst-case scenarios, but on the other hand, I don’t think it’s going to have a major day-to-day impact on forecasts.... We’re going to have to, for a while, deal with a bit more uncertainty than we’ve come to [expect].”
Anyway, I wrote the first few paragraphs of this post and then at 1:51 p.m. two weeks ago Tuesday, EARTHQUAKE! Then there was Hurricane Irene. Then flooding. So now, where was I?
Let me begin by providing a little context on where weather satellites came from and how they fit into the overall scheme of today's meteorological enterprise. As Irene proved, we have come a long way in making good forecasts, but can still struggle sometimes predicting tomorrow’s weather (not to mention the weather and climate our future generations will expect).
Forecasting the weather is now a solid scientific process, not the part-art form it was 50 or 100 years ago. Forecasting has been likened to a three-legged stool of data, physical science and computer science. The three legs must be about equal for the forecast to stand. Too little data, even with huge supercomputers, won’t produce an accurate forecast. Conversely, even if we have all the data in the world, we can't tell whether tomorrow will be sunny or overcast if we lack an understanding of how clouds form.
Since the first one was launched by the United States in 1959, weather satellites have become a very critical part of bolstering the “data” leg of a forecast. Look at this historic television image taken by Tiros-1, which achieved orbit in 1960 and is considered to be the first truly successful weather satellite:
And here is the GOES satellite image of the western hemisphere as Hurricane Irene churned our way:
It's easy to see the tremendous progress we have made. But the GOES satellite did not replace TIROS overnight, and that’s where we have run into some problems.
It takes years to develop, test and launch new meteorological satellites and their critical sensors, from high-resolution cameras to remote-sounding instruments that can measure ocean temperatures and moisture in the air from hundreds of miles away. Years ago, it seemed like a good idea to combine the needs of various government agencies for a new generation of polar orbiting satellites as a way of making our tax dollars work better. The program that came out of this idea in 1994 was called the National Polar-orbiting Operational Environmental Satellite System, or NPOESS.
For a number of reasons and despite the hard work of dedicated leaders, the NPOESS program eventually ran over budget by billions of dollars. Meanwhile, the time needed to replace old polar satellites grew shorter. So last year the program was split, and to meet the needs of the National Weather Service, the National Oceanic and Atmospheric Administration and NASA, the polar-satellite program was renamed the Joint Polar Satellite System. But time and money are tight and the development of this new-generation of weather satellite was slowed:
Then, this July, Congress proposed a major cut in funding that would probably mean the nation’s weather services would have only one polar satellite for perhaps a year rather than the two satellites we now depend on. That brought out headlines and letters of concerns from leading scientific professional societies. NOAA ran some simulations of what the atmospheric “models,” the core of forecasting, would show using only one polar satellite.
The agency used the famous Washington “Snowmageddon” of Feb. 2010 as an example. This is what the forecast looked like five days before with all the needed data from one satellite compared with the data from two satellites. Lots of changes compared to the actual snowfall – it's sort of like having meteorologists forecast with one hand tied behind their backs:
Now, what does this have to do with my quote in the Washington Post? As Paul Harvey would say, here's “the rest of the story.”
I thought and still think that my longtime friends and colleagues at NOAA and NWS using “Snowmageddon” as an example of the risks of losing satellite data (meteorological, not political, risks) is not necessarily the whole story. Why? Because predicting a snowstorm isn't the same thing as communicating about a snowstorm, and one is just as crucial as the other.
Take the Jan. 26, 2011 "Thundersnow" storm. The objective weather forecast done with that system, as well as with Snowmageddon, was excellent. Unbelievably detailed, correct physical forecasts of what would happen in the atmosphere days and especially 24 hours ahead. It was a terrific scientific achievement!
But what happened? Bad decisions were still made and in the case of the terrific Thundersnow forecast, the result was the worst traffic nightmare in Washington commuter history. This past January we correctly forecast the birth of a very unusual weather event... and ended up with Godzilla.
As Roger Pielke Jr. (who pens a must-read blog) and others have correctly pointed out, weather forecasting is not a single delivery of a forecast. It is an “end-to-end event.” First, there is the gathering of the data (for sure we DO NEED two polar satellites, not one) to generate the simulations we need to make a very accurate forecast. Then there must be the communication of the forecast to decision-makers including you, the public and finally the correct weather related decision is made.
I as a “media” meteorologist (never thought of myself as a digital entity until recently) have to effectively communicate the weather forecast to you so you will have an idea of what I know, and together we will make the same weather-related decision tomorrow or next week. If we fail to effectively communicate a perfect forecast then we are inviting poor decisions or, even worse, life-threatening decisions. We and I have failed.
We do have to better communicate. The last time I checked, the National Weather Service still had only a few social scientists on staff. I grew up, as many meteorologists, as a weather nerd. I love science, but that does not mean I still know how best to communicate what I know (I do try, for what it's worth). We have the capability now to issue a weather forecast much better than a one liner such as “Saturday rain or snow.” But that is part of the legacy of communications systems within a terrific scientific organization that serves our county every day with life saving information. Our National Weather Service.
The forecasting problem is not solved. The forecast track of Irene was terrific and communication was excellent. The forecast of the winds and coastal impacts was just OK and the forecast of inland flooding needed improvement. How and what we as a forecasting/communication/decision-making enterprise did during Irene also needed improvement. And the addition of another or five more satellites will not help that, if we still don’t help everyone make the right decision.
I believe that even a once-in-a-lifetime experience such as the terrible outbreak of EF-5 tornadoes this spring and tragic flooding in Vermont can be better communicated and prepared for by everyone if we devote resources to the still-weak links of communications for best decision-making. We are making progress through programs such as WAS*IS at the National Center for Atmospheric Research (NCAR) and other institutions. Some of the most satisfying advances in the application of the science I have loved for all my life are no doubt yet to come.
We spend about 15 cents per person per week for our nation's weather services. That's 15 cents per week for all the satellites, computers, weather instruments, meteorologists, research, hurricane hunter flights... everything. If we add one extra penny per person per week to study and learn how we can better communicate what we as scientists know so you can make the best decision the next time there is a Snowmageddon or an Irene, would you think that was wasted money?
Even at 16 cents per person per week, our nation's weather services (protecting life and property of its citizens is a fundamental purpose of government, isn't it?) is one of the best bargains for we taxpayers.