Weather Research

Soil moisture, snowpack data could help predict 'flash droughts'

BOULDER, Colo. — New research suggests that "flash droughts" — like the one that unexpectedly gripped the Southern Rockies and Midwest in the summer of 2012 — could be predicted months in advance using soil moisture and snowpack data. Researchers at the National Center for Atmospheric Research (NCAR) analyzed the conditions leading up to the 2012 drought, which ultimately caused $30 billion in economic losses, looking for any warning signs that a drought was on the way. In a study funded by the National Science Foundation and published in the Journal of Geophysical Research-Atmospheres, the scientists find that observations of snowmelt and soil moisture could have predicted the ensuing drought up to four months in advance."The 2012 drought over the Midwest was one of the most severe and extensive U.S. droughts since the 1930s Dust Bowl, but it was also extremely challenging to predict," said Debasish PaiMazumder, lead author of the study. "This study demonstrated the potential to improve seasonal drought outlooks in the future, giving farmers, water planners, and others more time to prepare."The official U.S. Drought Monitor issued on Aug. 21, 2012. The map shows the exceptionally severe drought across the middle of the country. Just three months before, drought forecasts failed to predict that a drought was on the way. Click to enlarge. (Image courtesy National Drought Mitigation Center.)Seasonal drought forecasts issued in May 2012 for the upcoming summer did not foresee a drought forming in the country's midsection. But by the end of August, a drought that had started in the Southern Rockies had spread across the Midwest, parching Oklahoma, Kansas, Nebraska, and Missouri.These flash droughts — which form and intensify rapidly — can catch forecasters off guard because they are not preceded by any large-scale climate patterns that could act as a warning signal. For example, one contributor to the recent California drought was a persistent high-pressure system parked off the west coast of Canada that deflected storms away from the state. Because forecasters could identify the high-pressure system, they could also accurately predict fewer storms and a worsening of the drought.Previous research has shown that looking at soil moisture alone could improve the lead-time of drought predictions by one to two months. PaiMazumder and NCAR colleague James Done were interested in whether they could extend this further by adding snowpack into the equation.“Advance knowledge of a drought even a month or two ahead of time can greatly minimize the effects on society,” said Anjuli Bamzai, program director in NSF’s Division of Atmospheric and Geospace Sciences, which funded the research.  “This study highlights the role of snowpack and soil moisture conditions in predicting the sudden onset of drought.”To explore the physical connections among snowpack, soil moisture, and drought, the researchers analyzed data collected between 1980-2012. To supplement those observations, they also explored the physical connections in a new NCAR-based community Weather Research and Forecasting (WRF) model dataset comprising 24 simulations of the period 1990-2000 and 2012. Because each simulation was run with small tweaks to the way the model represents atmospheric physics, the result was a broad look at different climate scenarios that could have plausibly unfolded during the study period."The model helped us get a handle on how robust the relationships between snowpack, soil moisture, and drought are," Done said. "The stronger the relationship, the better a predictor is."While observations of snowpack and soil moisture could have helped predict the 2012 drought, the method does not replace other drought prediction measures that identify large-scale phenomena that frequently lead to drought conditions."This is another ingredient that could be used when making seasonal drought forecasts," Done said. "But it's not the only ingredient, and for many droughts that are tied to large-scale precursors, it may not be the most important one." About the articleTitle: Potential Predictability Sources of the 2012 US Drought in Observations and a Regional Model EnsembleAuthors: Debasish PaiMazumder and James DoneJournal: Journal of Geophysical Research – Atmospheres, DOI: 10.1002/2016JD025322Writer:Laura Snider, Senior Science Writer and Public Information Officer   

Advanced computer model focuses on Hurricane Matthew

Oct. 6, 2016 | As Hurricane Matthew churns toward the southeastern U.S. coast, scientists at the National Center for Atmospheric Research (NCAR) are testing an advanced research computer model to see how well it can predict the powerful storm's track and intensity.The Model for Prediction Across Scales (MPAS) uses an innovative software approach that allows scientists to focus on regional conditions while still capturing far-flung atmospheric processes that can influence the storm in question. This is a contrast to the forecast models typically used to track hurricanes today, which cannot simultaneously capture both global and local atmospheric processes.The experimental MPAS model simulates Hurricane Matthew hitting the Southeast. To see a range of model output, visit the MPAS tropical cyclone website. MPAS is able to do both because it uses a flexible mesh that allows it to zoom into higher resolution in some areas — over hurricane breeding grounds, for example — while zooming out over the rest of Earth. This ability to vary resolution across the globe requires a small fraction of the computer power needed to have high resolution everywhere.By testing MPAS during hurricane season, the research team can determine the adjustments that need to be made to the model while gaining insights into how to improve hurricane forecasting in the future."This is an experimental effort," said Chris Davis, a senior scientist and director of NCAR's Mesoscale and Microscale Meteorology Laboratory. "We're doing this to see if we can find systematic biases in the model so we can improve simulations of the tropics in general and hurricanes in particular."Davis and the other members of the research team, including NCAR scientists David Ahijevych, Sang-Hun Park, Bill Skamarock, and Wei Wang, are running MPAS once a day on NCAR's Yellowstone supercomputer, inputting various ocean and atmospheric conditions to see how it performs. The work is supported by the National Science Foundation and the Korea Institute of Science and Technology Information.Even though they are just tests, Davis said the MPAS simulations are often comparable with official forecast models such as those run by the National Hurricane Center and the European Centre for Medium-Range Weather Forecasts. As Matthew was in its early stages, in fact, MPAS did a better job than other models in simulating the northward movement of the storm from the Caribbean Sea toward the Florida coast.The scientists will analyze how MPAS performed and share the results with colleagues in the meteorological community. It's a step in an ongoing research effort to better predict the formation and behavior of hurricanes."We run the model even when the tropics are quiet, but an event like Matthew gives us a special opportunity to see what contributes to errors in tropical cyclone prediction," Davis said. "While a major hurricane can have catastrophic impacts, we hope to learn from it and make computer models even better in the future."Funders:National Science FoundationKorea Institute of Science and Technology InformationWriter/contact:David Hosansky, Manager of Media Relations

UCAR congressional briefing highlights flood, drought prediction

WASHINGTON — The nation is poised to make major advances in "water intelligence" with more detailed forecasts of floods, streamflow, and potential drought conditions, a panel of experts said at a congressional briefing today.The briefing, sponsored by the University Corporation for Atmospheric Research (UCAR), highlighted the new National Water Model, a comprehensive system for forecasting water resources from coast to coast. The technology underpinning the model, launched last month by the National Oceanic and Atmospheric Administration (NOAA), was developed by the National Center for Atmospheric Research (NCAR) and its collaborators at universities, the National Science Foundation and other federal agencies, and the private sector."The new forecast model is really a quantum leap forward and will help safeguard Americans from major floods and other precipitation events," said UCAR President Antonio J. Busalacchi, who introduced the panel. "It bridges the gap between research and operations, generating real-time forecasts to help vulnerable communities and protect lives and property."UCAR manages NCAR on behalf of the National Science Foundation."Through a series of partnerships, it's possible to provide consistent, high-resolution, integrated water analyses, predictions, and data to address critical unmet information and service gaps," said Edward Clark, director of the Geo-Intelligence Office of Water Prediction at the NOAA National Water Center.Scientists generated this inundation forecast during Houston-area flooding earlier this year in a demonstration of  advanced computer modeling technology. (©UCAR. Image by David Gochis, NCAR. This image is freely available for media & nonprofit use.)Unlike past streamflow models, which provided forecasts every few hours and only for specific points along major river systems, the new system continuously simulates conditions at 2.7 million locations along rivers, streams, and their tributaries across the contiguous United States. It paves the way for the biggest improvement in flood forecasting in the nation's history."The National Water Model provides a different way of thinking about continental hydrology by providing a view of a connected plumbing network from the mountains to the ocean," said panelist Richard Hooper, executive director of the Consortium of Universities for the Advancement of Hydrologic Science (CUAHSI). "Previously, hydrologists had considered river basins as discrete units rather than this river-continuum approach. This change in view opens up new areas of research that will improve our ability to predict not just floods but other aspects of water resources, including water quality and the impacts of droughts."Thanks to ongoing research, the National Water Model is expected to provide increasingly detailed street-level forecasts, inundation maps, and additional features such as water quality forecasts. Scientists are working on incorporating more processes, such as soil saturation and the amount of water drawn up by vegetation."By dramatically increasing the geographic coverage as well as the lead times for forecasts, the National Water Model is ushering in a new era in flood and flash flood forecasting," said John McHenry, chief scientist of advanced meteorological systems for Baron Services. "Business, industry, and the general public will benefit through reduction in lost lives and property."The panelists emphasized the importance of water resources to the major sectors of the U.S. economy. They warned that the nation is facing myriad water-related challenges ranging from growing demand to increasingly costly floods and droughts. Meeting those challenges will require continued coordination among research organizations, universities, the private sector, and federal, state, and local agencies."Beyond developing a new computer model, we're building a community by sharing resources, tools, and ideas," said NCAR scientist David Gochis. "The scientists are engaging with practitioners and decision makers to make the system as usable as possible."The development team at NCAR worked with scientists at NOAA, the U.S. Geological Survey, and universities to adapt WRF-Hydro to serve as the first version of the National Water Model.The panelists also discussed the need for better water intelligence among diverse communities across the country. For example, Ryan Emanuel, associate professor at North Carolina State University's Department of Forestry and Environmental Resources, noted that indigenous tribes across the nation are particularly vulnerable to drought and flooding for a range of cultural, historical, and economic reasons."Indigenous peoples across the United States are diverse, but one common theme is that water is sacred," said Emanuel, a member of the Lumbee Tribe of North Carolina. "It's not only critical for life, but it is life itself. Beyond the tools, the models, and the management lies the knowledge of the original inhabitants of this nation that water binds us all to a common fate."The event is the latest in a series of UCAR congressional briefings about critical topics in the Earth system sciences. Past briefings have focused on predicting space weather, aviation weather safety, the state of the Arctic, hurricane prediction, and potential impacts of El Niño.

NCAR|UCAR hurricane experts available to explain storm behavior, impacts

BOULDER, Colo. — Scientists at the National Center for Atmospheric Research (NCAR) and its managing organization, the University Corporation for Atmospheric Research (UCAR), use advanced computer models and observations to study how tropical storms behave and their impacts on society. Our hurricane experts are available to explain:Why hurricanes and tropical storms form and what causes the behavior of these powerful storms;How we can better predict the possible impacts of hurricanes, including flooding and subsequent spread of disease-bearing mosquitoes;How climate change may be impacting hurricanes and what can we expect in the future; andHow well hurricane forecasts are communicated and how communication can be improved. Hurricane formation, intensity, and trackChristopher Davis, director, NCAR Mesoscale and Microscale Meteorology Laboratory,, 303-497-8990. Davis studies the weather systems that lead to hurricanes and other heavy rainfall events. His expertise includes hurricane prediction and how computer models can be improved to better forecast storms.Jeff Weber, UCAR meteorologist,, 303-497-8676. An expert on hurricanes and severe weather in general, Weber closely monitors the behavior of individual storms and the larger atmospheric and oceanic conditions that influence them. FloodingMatthew Kelsch, UCAR hydrometeorologist,, 303-497-8309. Kelsch has studied some of the biggest U.S. flood events connected to hurricanes and tropical storms, and he trains scientists from around the world on hydrology topics.David Gochis, NCAR scientist,, 303-497-2809. An expert in hydrometeorology, Gochis studies the causes of floods and how to better predict them. Communicating forecastsRebecca Morss, NCAR Scientist,, 303-497-8172. Morss studies how hurricane and flash flood risks and evacuation plans can be better communicated to the public.Julie Demuth, NCAR scientist,, 303-497-8112. Demuth studies how to better communicate forecasts and warnings to the public, and how to create more useful weather hazard products and tools for forecasters. Economic and societal ImpactsJeff Lazo, NCAR scientist,, 303-497-2857. Lazo leads NCAR's societal impacts program. He has researched the value of improving hurricane forecast accuracy and the potential need for a new National Weather Service storm surge warning product. Climate changeKevin Trenberth, NCAR senior scientist,, 303-497-1318. Trenberth has been in the forefront of scientists examining the potential influence of climate change on the intensity of tropical storms and hurricanes.Greg Holland, NCAR senior scientist, Holland, an expert on possible links between global warming and tropical cyclone activity, is helping to lead a program at NCAR to develop tools that will enhance society's resilience to extreme weather.James Done, NCAR scientist,, 303-497-8209. Done recently helped develop a new index to quantify a hurricane's ability to cause destruction. The Cyclone Damage Potential (CDP) index rates storms on a scale of 1 to 10. It can also be used to examine how the potential for cyclones to cause damage may change in the future as the climate warms. DiseaseMary Hayden, NCAR scientist,, 303-497-8116. Hayden investigates the link between climate and disease, and she collaborates with universities and health officials to reduce societal risk and vulnerability to disease.Andrew Monaghan, NCAR scientist,, 303-497-8424. Monaghan uses computer modeling to identify the potential risks of diseases such as the Zika virus and dengue fever due to climate and other factors.

NSF/NCAR research plane assisting with U.S. hurricane forecasts

BOULDER, Colo. — As the peak of hurricane season approaches, U.S. forecasters are deploying a high-altitude research aircraft operated by the National Center for Atmospheric Research (NCAR) to fly over and around storms to take critical observations.The NSF/NCAR Gulfstream-V readies for takeoff on a mission to study tropical storms. (©UCAR. Photo by Carlye Calvin. This image is freely available for media & nonprofit use.) The deployment this week of the Gulfstream-V (G-V) aircraft is the result of a partnership between the National Science Foundation (NSF), which owns the plane, and the National Oceanic and Atmospheric Administration (NOAA), which issues forecasts. The NSF/NCAR G-V will take to the skies to support hurricane forecasts through October 12, while NOAA’s Gulfstream-IV (G-IV) undergoes unscheduled maintenance."It's critical to have detailed measurements of the atmosphere around a hurricane in order to ensure that forecasts are as accurate as possible," said Antonio (Tony) J. Busalacchi, president of the University Corporation for Atmospheric Research, which manages NCAR on behalf of NSF. "NCAR and its research partners have a proven track record of improving predictions of dangerous storms. Consistent with our role of managing NCAR, we take very seriously our ability and responsibility to share our advanced resources in support of NOAA's mission to protect life and property.""NSF is pleased that NCAR, using the G-V, is able to assist in this potentially lifesaving activity," said Roger Wakimoto, assistant director of the NSF Directorate for Geosciences. "The data gathered will help refine future hurricane forecasts.”Outfitted for critical observationsThe NSF/NCAR G-V can fly at high altitudes and deploy the same specialized sensors as the NOAA G-IV. These sensors take critical observations of atmospheric conditions for the NOAA National Hurricane Center.Studies show that such observations improve hurricane track forecasts in the U.S. global weather model (called the GFS) by about 15 percent during the 24 to 48 hours before landfall. Research also demonstrates that these data increase the accuracy of hurricane intensity forecasts.To take the observations, the NSF/NCAR G-V has been outfitted with the Airborne Vertical Atmospheric Profiling System (AVAPS). The system releases parachute-borne sensors, known as GPS dropsondes, that measure ambient temperature, pressure, humidity, wind speed, and wind direction at different altitudes as they fall through the atmosphere. Dropsondes were first developed at NCAR in the 1970s with NSF funding and have since been regularly updated. NOAA was an early adopter of the dropsondes for hurricane surveillance missions and research, and the development of the AVAPS system design in the 1990s was motivated in part by the capabilities of the NOAA G-IV.The NSF/NCAR G-V, which is available for flights over both the Atlantic and Pacific, will fly above a hurricane or other major storm at altitudes of up to 45,000 feet, as well as around the storm's edges. Its dropsonde launch system and software is similar to that of the NOAA G-IV.NCAR pilots will guide the aircraft on pre-planned flight tracks, dropping sondes approximately every 15 minutes. Data from the sondes will be processed by a NOAA technician onboard the plane, then sent to the Global Telecommunications System for immediate inclusion in hurricane forecast models."It is a special privilege for us to be able to help out our colleagues at NOAA by deploying the NSF/NCAR G-V in the hurricane surveillance missions this season," said Vanda Grubišić, director of NCAR's Earth Observing Laboratory, which operates the G-V. "Our Research Aviation Facility crews look forward to working with their NOAA colleagues and collecting important data in support of their mission."

Solar energy gets boost from new forecasting system

BOULDER, Colo. — A cutting edge forecasting system developed by a national team of scientists offers the potential to save the solar energy industry hundreds of millions of dollars through improved forecasts of the atmosphere.The new system, known as Sun4CastTM, has been in development for three years by the National Center for Atmospheric Research (NCAR) in collaboration with government labs, universities, utilities, and commercial firms across the country. Funded by the U.S. Department of Energy SunShot Initiative, the system greatly improves predictions of clouds and other atmospheric conditions that influence the amount of energy generated by solar arrays.After testing Sun4Cast at multiple sites, the research team has determined that it can be up to 50 percent more accurate than current solar power forecasts. This improved accuracy will enable utilities to deploy solar energy more reliably and inexpensively, reducing the need to purchase energy on the spot market. The amount of energy gathered by solar panels — such as these in Colorado's San Luis Valley — is influenced by factors including the position and types of clouds, the amount of snow on the ground, and relative humidity. The new Sun4Cast system greatly improves solar irradiance predictions, enabling utilities to deploy solar energy more reliably and inexpensively. (©UCAR. Photo by Sue Ellen Haupt, NCAR. This image is freely available for media & nonprofit use.)As a result, utilities across the United States may be able to save an estimated $455 million through 2040 as they use more solar energy, according to an analysis by NCAR economist Jeffrey Lazo.NCAR, which does not provide operational forecasts, makes the technology available so it can be adapted by utilities or private forecasting companies. The research is being highlighted in more than 20 peer-reviewed papers."These results can help enable the nation's expanding use of solar energy," said Sue Ellen Haupt, director of NCAR’s Weather Systems and Assessment Program, who led the research team. "More accurate predictions are vital for making solar energy more reliable and cost effective."The work builds on NCAR’s expertise in highly detailed atmospheric prediction, including the design of an advanced wind energy forecasting system."This type of research and development is important because it contributes to the reduction in costs for solar and wind energy and makes it easier for utilities to integrate renewables into the electrical grid," said William Mahoney, Deputy Director of NCAR's Research Applications Laboratory. "When it comes to balancing demand for power with supply, it's vital to be able to predict sources of energy as accurately as possible."Xcel Energy is already beginning to use the system to forecast conditions at several of its main solar facilities.“Our previous experience with the National Center for Atmospheric Research in developing a wind forecasting system has saved millions of dollars and has been highly beneficial for our customers," said Drake Bartlett, senior trading analyst for Xcel Energy – Colorado. "It is our sincere hope and belief that we will see positive atmospheric forecasting results for predicting solar generation as well, again to the benefit of our Xcel Energy customers."Energy forecasts out to 72 hoursUsing a combination of advanced computer models, atmospheric observations, and artificial intelligence techniques, Sun4Cast provides 0- to 6-hour nowcasts of solar irradiance and the resulting power production for specific solar facilities at 15-minute intervals. This enables utilities to continuously anticipate the amount of available solar energy.In addition, forecasts extend out to 72 hours, allowing utility officials to make decisions in advance for balancing solar with other sources of energy.Solar irradiance is notoriously difficult to predict. It is affected not just by the locations and types of clouds, but also a myriad of other atmospheric conditions, such as the amount of dust and other particles in the air, relative humidity, and air pollution. Further complicating the forecast, freshly fallen snow, nearby steep mountainsides, or even passing cumulus clouds can reflect sunlight in a way that can increase the amount of energy produced by solar panels.To design a system to forecast solar energy output, NCAR and its partners drew on an array of observing instruments, including satellites, radars, and sky imagers; specialized software; and mathematical and artificial intelligence techniques. Central to Sun4Cast is a new computer model of the atmosphere that simulates solar irradiance based on meteorological conditions. Called WRF-SolarTM, the model is derived from the NCAR-based Weather Research and Forecasting (WRF) model, which is widely used by meteorological agencies worldwide.The team tested the system in geographically diverse areas, including Long Island, New York; the Colorado mountains; and coastal California."We have to provide utilities with confidence that the system maintains a high degree of accuracy year-round in very different types of terrain," said Branko Kosovic, NCAR Program Manager for Renewable Energy.In addition to aiding the solar power industry, the work can also improve weather forecasting in general because of improved cloud prediction.NCAR's numerous partners on the project in the public and private sectors included:Government labs: National Renewable Energy Laboratory, Brookhaven National Laboratory, the National Oceanic and Atmospheric Administration’s Earth System Research Laboratory, and other NOAA facilities; Universities: The Pennsylvania State University, Colorado State University, University of Hawaii, and University of Washington; Utilities: Long Island Power Authority, New York Power Authority, Public Service Company of Colorado, Sacramento Municipal Utility District (SMUD), Southern California Edison, and the Hawaiian Electric Company; Independent system operators: New York ISO, Xcel Energy, SMUD, California ISO, and Hawaiian Electric; and Commercial forecast providers: Schneider Electric, Atmospheric and Environmental Research, Global Weather Corporation, MDA Information Systems, and Solar Consulting Services.Computing time was provided by the New York State Department of Economic Development's Division of Science, Technology and Innovation on an IBM Blue Gene supercomputer at Brookhaven National Laboratory. Researchers also performed computing at the NCAR-Wyoming Supercomputing Center and the DOE National Energy Research Scientific Computing Center.About the SunShot InitiativeThe U.S. Department of Energy SunShot Initiative is a collaborative national effort that aggressively drives innovation to make solar energy fully cost-competitive with traditional energy sources before the end of the decade. Through SunShot, the Energy Department supports efforts by private companies, universities, and national laboratories to drive down the cost of solar electricity to $0.06 per kilowatt-hour.

US taps NCAR technology for new water resources forecasts

BOULDER, Colo. — As the National Oceanic and Atmospheric Administration (NOAA) this month launches a comprehensive system for forecasting water resources in the United States, it is turning to technology developed by the National Center for Atmospheric Research (NCAR) and its university and agency collaborators.WRF-Hydro, a powerful NCAR-based computer model, is the first nationwide operational system to provide continuous predictions of water levels and potential flooding in rivers and streams from coast to coast. NOAA's new Office of Water Prediction selected it last year as the core of the agency's new National Water Model."WRF-Hydro gives us a continuous picture of all of the waterways in the contiguous United States," said NCAR scientist David Gochis, who helped lead its development. "By generating detailed forecast guidance that is hours to weeks ahead, it will help officials make more informed decisions about reservoir levels and river navigation, as well as alerting them to dangerous events like flash floods."WRF-Hydro (WRF stands for Weather Research and Forecasting) is part of a major Office of Water Prediction initiative to bolster U.S. capabilities in predicting and managing water resources. By teaming with NCAR and the research community, NOAA's National Water Center is developing a new national water intelligence capability, enabling better impacts-based forecasts for management and decision making.The new WRF-Hydro computer model simulates streams and other aspects of the hydrologic system in far more detail than previously possible. (Image by NOAA Office of Water Prediction.) Unlike past streamflow models, which provided forecasts every few hours and only for specific points along major river systems, WRF-Hydro provides continuous forecasts of millions of points along rivers, streams, and their tributaries across the contiguous United States. To accomplish this, it simulates the entire hydrologic system — including snowpack, soil moisture, local ponded water, and evapotranspiration — and rapidly generates output on some of the nation's most powerful supercomputers.WRF-Hydro was developed in collaboration with NOAA and university and agency scientists through the Consortium of Universities for the Advancement of Hydrologic Science, the U.S. Geological Survey, Israel Hydrologic Service, and Baron Advanced Meteorological Services. Funding came from NOAA, NASA, and the National Science Foundation, which is NCAR's sponsor."WRF-Hydro is a perfect example of the transition from research to operations," said Antonio (Tony) J. Busalacchi, president of the University Corporation for Atmospheric Research, which manages NCAR on behalf of the National Science Foundation (NSF). "It builds on the NSF investment in basic research in partnership with other agencies, helps to accelerate collaboration with the larger research community, and culminates in support of a mission agency such as NOAA. The use of WRF-Hydro in an operational setting will also allow for feedback from operations to research. In the end this is a win-win situation for all parties involved, chief among them the U.S. taxpayers.""Through our partnership with NCAR and the academic and federal water community, we are bringing the state of the science in water forecasting and prediction to bear operationally," said Thomas Graziano, director of NOAA’s new Office of Water Prediction at the National Weather Service.Filling in the water pictureThe continental United States has a vast network of rivers and streams, from major navigable waterways such as the Mississippi and Columbia to the remote mountain brooks flowing from the high Adirondacks into the Hudson River. The levels and flow rates of these watercourses have far-reaching implications for water availability, water quality, and public safety.Until now, however, it has not been possible to predict conditions at all points in the nation's waterways. Instead, computer models have produced a limited picture by incorporating observations from about 4,000 gauges, generally on the country's bigger rivers. Smaller streams and channels are largely left out of these forecast models, and stretches of major rivers for tens of miles are often not predicted — meaning that schools, bridges, and even entire towns can be vulnerable to unexpected changes in river levels.To fill in the picture, NCAR scientists have worked for the past several years with their colleagues within NOAA, other federal agencies, and universities to combine a range of atmospheric, hydrologic, and soil data into a single forecasting system.The resulting National Water Model, based on WRF-Hydro, simulates current and future conditions on rivers and streams along points two miles apart across the contiguous United States. Along with an hourly analysis of current hydrologic conditions, the National Water Model generates three predictions: an hourly 0- to 15-hour short-range forecast, a daily 0- to 10-day medium-range forecast, and a daily 0- to 30-day long-range water resource forecast.The National Water Model predictions using WRF-Hydro offer a wide array of benefits for society. They will help local, state, and federal officials better manage reservoirs, improve navigation along major rivers, plan for droughts, anticipate water quality problems caused by lower flows, and monitor ecosystems for issues such as whether conditions are favorable for fish spawning. By providing a national view, this will also help the Federal Emergency Management Agency deploy resources more effectively in cases of simultaneous emergencies, such as a hurricane in the Gulf Coast and flooding in California."We've never had such a comprehensive system before," Gochis said. "In some ways, the value of this is a blank page yet to be written."A broad spectrum of observationsWRF-Hydro is a powerful forecasting system that incorporates advanced meteorological and streamflow observations, including data from nearly 8,000 U.S. Geological Survey streamflow gauges across the country. Using advanced mathematical techniques, the model then simulates current and future conditions for millions of points on every significant river, steam, tributary, and catchment in the United States.In time, scientists will add additional observations to the model, including snowpack conditions, lake and reservoir levels, subsurface flows, soil moisture, and land-atmosphere interactions such as evapotranspiration, the process by which water in soil, plants, and other land surfaces evaporates into the atmosphere.Scientists over the last year have demonstrated the accuracy of WRF-Hydro by comparing its simulations to observations of streamflow, snowpack, and other variables. They will continue to assess and expand the system as the National Water Model begins operational forecasts.NCAR scientists maintain and update the open-source code of WRF-Hydro, which is available to the academic community and others. WRF-Hydro is widely used by researchers, both to better understand water resources and floods in the United States and other countries such as Norway, Germany, Romania, Turkey, and Israel, and to project the possible impacts of climate change."At any point in time, forecasts from the new National Water Model have the potential to impact 300 million people," Gochis said. "What NOAA and its collaborator community are doing is trying to usher in a new era of bringing in better physics and better data into forecast models for improving situational awareness and hydrologic decision making."CollaboratorsBaron Advanced Meteorological Services Consortium of Universities for the Advancement of Hydrologic ScienceIsrael Hydrologic ServiceNational Center for Atmospheric ResearchNational Oceanic and Atmospheric AdministrationU.S. Geological SurveyFundersNational Science FoundationNational Aeronautics and Space AdministrationNational Oceanic and Atmospheric Administration

NCAR weather ensemble offers glimpse at forecasting's future

July 1, 2016 | Last spring, scientists at the National Center for Atmospheric Research (NCAR) flipped the switch on a first-of-its-kind weather forecasting system. For more than a year, NCAR's high-resolution, real-time ensemble forecasting system has been ingesting 50,000 to 70,000 observations every six hours and creating a whopping 90,000 weather maps each day.The system has become a favorite among professional forecasters and casual weather wonks: Typically more than 200 people check out the site each day with more than a thousand coming during major weather events. During this experimental period, the NCAR ensemble has also become a popular source of guidance within the National Weather Service, where it has already been referenced several hundred times by forecasters at more than 50 different offices. But perhaps more important, the data accumulated from running the system daily — and there is lots of it — is being used by researchers at universities across the country to study a range of topics, from predicting hail size to anticipating power outages for utilities."We wanted to demonstrate that a real-time system of this scale was feasible," said NCAR scientist Craig Schwartz. "But it's also a research project that can help the community learn more about the predictability of different kinds of weather events."Schwartz is a member of the team that designed and operates the system, along with NCAR colleagues Glen Romine, Ryan Sobash, and Kate Fossell.This animation shows the forecast for accumulated snowfall made by each of the NCAR ensemble's 10 members for the 48-hour period beginning on Jan. 22, 2016. In the run-up to the blizzard, which ultimately dropped more than 30 inches of snow on parts of the Mid-Atlantic, more than 1,000 people visited the NCAR ensemble's website. (©UCAR. This animation is freely available for media & nonprofit use.)Testing a unique toolNCAR's high-resolution ensemble forecasting system is unique in the country for a couple of reasons, both of which are revealed in its name: It's an ensemble, and it's high resolution.Instead of producing a single forecast, the system produces an "ensemble" of 10 forecasts, each with slightly different (but equally likely) starting conditions. The degree to which the forecasts look the same or different tells scientists something about the probability that a weather event, like rain, hail, or wind, will actually occur.By comparing the actual outcomes to the forecasted probabilities, scientists can study the predictability of particular weather events under different circumstances. The forecasting system's high resolution (the grid points are just 3 kilometers apart) allows it to simulate small-scale weather phenomena, like the creation of individual storms from convection — the process of moist, warm air rising and then condensing into clouds. The combination of fine grid spacing and ensemble predictions in the NCAR system offers a sneak peek at what the future of weather forecasting might look like, and weather researchers across the country have noticed. Cliff Mass, a professor of atmospheric sciences at the University of Washington whose specialty is forecasting, said: "It's extremely important for the United States to have a convection-allowing ensemble system to push our forecasting capabilities forward. We were delighted that NCAR demonstrated that this could be done."'The cat's meow'The treasure trove of accruing weather data generated by running the NCAR ensemble is already being used by researchers both at NCAR and in the broader community. Jim Steenburgh, for instance, is a researcher at the University of Utah who is using the system to understand the predictability of mountain snowstorms."NCAR's ensemble not only permits the 'formation' of clouds, it can also capture the topography of the western United States," he said. "The mountains control the weather to some degree, so you need to be able to resolve the mountains' effects on precipitation."Steenburgh has also been using the ensemble with his students. "We’re teaching the next generation of weather forecasters," he said. "In the future, these high-resolution ensemble forecasts will be the tools they need to use, and this gives them early, hands-on experience." Like Steenburgh, Lance Bosart, an atmospheric researcher at the University of Albany, State University of New York, has used the ensemble both in his own research — studying the variability of convective events — and with his students. He said having 10 members in the ensemble forecast helps students easily see the great spread of possibilities, and the visual emphasis of the user interface makes it easy for students to absorb the information."What makes it an invaluable tool is the graphical display," he said. "It's visually compelling. You don't have to take a lot of time to explain what you're looking at; you can get right into explaining the science. I like to say it's the cat's meow." Setting an exampleThe NCAR ensemble is also enabling the researchers running it to further their own research. "We're collecting statistics on the misfit between the model predictions and observations and then we're trying to use that to improve our model physics," Romine said. The ensemble project is also teaching the team about the strengths and weaknesses of the way they've chosen to kick off, or "initialize," each of the ensemble members. "The NCAR ensemble happens to produce a pretty good forecast, but we realize there are some shortcomings," Schwartz said. "For example, if we were trying to make the best forecast in the world, we would probably not be initializing the model the way we are. But then we wouldn’t learn as much from a research perspective." The NCAR ensemble began as a yearlong trial, but the project is continuing to run for now. The team would like to keep the system online until next summer, but they don't yet have the computing resources they need to run it past September.If the system does continue to run, the researchers who are using it say there's still more that they and their students can learn from the project. And if not, there's loads of data already collected that are still waiting to be mined. In any case, Mass says NCAR's ensemble has been a valuable project. "It set a really good example for the nation," he said.Community members interested in collaborating or helping support the NCAR ensemble project are encouraged to contact the team at Snider, Senior Science Writer and Public Information Officer 

Sizing up cyclones

UPDATE: 2016 SEASONAL FORECASTThe NCAR-based Engineering for Climate Extremes Partnership (ECEP) has issued its first seasonal forecast using the Cyclone Damage Potential index. The forecast is for a hurricane season with higher-than-average potential to cause damage. This year's forecasted seasonal CDP is 5.7, compared to an average seasonal CDP of 3.7 for the years 1981 - 2010. For more details, visit the ECEP website.May 18, 2016 |In early July 2005, Hurricane Dennis, a Category 3 storm on the Saffir-Simpson Hurricane Wind Scale, was bearing down on the Gulf Coast. Anyone paying attention would have been forgiven for having a foreboding sense of déjà vu. Just 10 months earlier, another Category 3 storm, Hurricane Ivan, had followed a strikingly similar track, making landfall just west of Gulf Shores, Alabama. Ivan ravaged the region, ultimately causing an estimated $18.8 billion in damages. But Dennis, despite roaring ashore in practically the same neighborhood, caused only $2.5 billion in damages—less than one-seventh that caused by Ivan.The fact that two Category 3 hurricanes making similar landfall less than one year apart had such different impacts illustrates a weakness in the Saffir-Simpson scale, the system most commonly used by weather forecasters to categorize hurricane risk.Scientists at the National Center for Atmospheric Research (NCAR)—in collaboration with global insurance broker Willis—have developed a new index that they suspect can do a better job of quantifying a hurricane's ability to cause destruction. The Cyclone Damage Potential index (CDP) rates storms on a scale of 1 to 10, with 10 being the greatest potential for destruction.A prototype for an app that will allow risk managers to easily use the CDP to identify local damage potential is already available and will be among the first tools included in the beta version of the NCAR-based Global Risk, Resilience, and Impacts Toolbox when it is released later this year.Infrared satellite imagery of Hurricane Ivan (left) and Hurricane Dennis (right). Both storms were rated Category 3, both made landfall in almost the same area, and yet they caused vastly different amounts of damage. Click to enlarge. (Images courtesy NOAA.)Moving beyond wind speedOn the frequently used Saffir-Simpson Hurricane Wind Scale, hurricanes are placed in one of five categories, based on their sustained wind speeds. On the low end, Category 1 storms have sustained winds between 74–95 mph and are expected to cause "some damage." On the high end, Category 5 storms have sustained winds of 157 mph or higher and are expected to cause "catastrophic damage."Because the Saffir-Simpson scale relies solely on sustained wind speeds, it does not take into account all the characteristics of a storm linked to its destructive power."Hurricane wind damage is driven by more than simply wind speed," said James Done, one of three NCAR scientists who worked on the CDP. "The hurricane's size and forward speed also are important. A large, slowly moving hurricane that repeatedly batters an area with high winds can cause greater total damage than a smaller, faster hurricane that blows quickly through a region."Damage caused to a marina in New Orleans by Hurricane Katrina. Katrina would have received a CDP rating of 6.6, compared to a 5.0 for Hurricane Ivan and a 2.4 for Hurricane Dennis. (Image courtesy NOAA. Click here for high resolution.)For example, the critical difference between Ivan and Dennis turned out to be hurricane size, according to a study of the storms by Done and Jeffrey Czajkowski at the University of Pennsylvania's Wharton Risk Management and Decision Processes Center.To create the CDP, the scientists incorporated hurricane size and forward speed into their index, along with sustained winds. To determine the relative importance of each, the team used hurricane damage statistics from the hundreds of offshore oil and gas facilities that pepper the northern Gulf of Mexico. Because facilities are spread more-or-less evenly across the region, their exposure to hurricanes is approximately the same. Damage differences from storm to storm can therefore be attributed to differences in the storms themselves. The CDP does not predict actual damage – which could vary markedly, depending on where (or if) a hurricane makes landfall – but instead predicts the storm's potential.When applying the CDP to past hurricanes, the index was able to discern a difference between Ivan, which would have scored 5.0 on the CDP prior to landfall, and the much smaller Dennis, which would have earned a 2.4. The CDP rating for Hurricane Katrina, which ultimately caused more than $80 billion in damages in 2005, would have been 6.6.“The value of the index is in comparing current storms with storms from the past," Done said. "For example, if a hurricane is approaching New Orleans, you can compare its CDP with Hurricane Katrina's CDP and get a fuller picture of how much damage the storm is likely to cause."The CDP project was led by NCAR scientist Greg Holland, along with NCAR colleagues Done, Ming Ge, and Willis collaborator Rowan Douglas.Dive deeperFrom today's storm to tomorrow's climateIn its original form, the CDP can be easily applied in real time to existing hurricanes. But Done also wanted to find a way to examine how hurricane damage might change in the future, especially as the climate warms. The question of how climate change may influence hurricanes has been difficult to answer, in part because global climate models are typically not able to "see" the small-scale details of individual storms. Though some scientists have run climate models at a resolution that is fine enough to study hurricane formation, the effort requires so much computing power that it hasn't been practical to replicate on a large scale.To skirt this problem, hurricane researchers have looked for links between hurricane activity and phenomena that climate models can see—for example, the sea surface temperatures of ocean basins."People have used large-scale variables to infer tropical cyclone activity for decades," Done said. "I've done a similar thing, but instead of predicting how many hurricanes will form, I’m predicting hurricane damage potential."To make this "climate" version of the CDP, Done – together with NCAR colleagues Debasish PaiMazumder and Erin Towler, and Indian Space Research Organization collaborator Chandra Kishtawal – looked for variables in the large-scale environment that could be correlated to the three variables used for the original CDP: sustained winds, size, and storm speed.The team found that "steering flow," the winds that would blow along a hurricane, is correlated with forward speed. They also found that relative sea surface temperature – the difference between temperatures in the Atlantic and Pacific ocean basins – is linked to seasonal average hurricane intensity and size. This is because relative sea surface temperatures affect wind speeds higher up in the atmosphere, which in turn affect hurricane formation. The result is an index that can spit out a damage potential rating for a season, a year, or even longer, without needing to predict how many individual storms might form. Such forecasts are of interest to large reinsurance companies, like Willis Re and others."This technique enables us to translate our climate model simulations into information about extreme events that’s critical for businesses and policy makers,” Done said.Writer/ContactLaura Snider, Senior Science Writer and Public information OfficerFundersResearch Partnership to Secure Energy for AmericaWillis Re CollaboratorsEngineering for Climate Extremes PartnershipWillis Re

A 3D window into a tornado

This simulation was created by NCAR scientist George Bryan to visualize what goes on inside a tornado. The animation is the "high swirl" version in a series that goes from low, to medium, to high. Click to enlarge. (Courtesy Goerge Bryan, NCAR. This image is freely available for media & nonprofit use.) May 17, 2016 | What's really going on inside a tornado? How fast are the strongest winds, and what are the chances that any given location will experience them when a tornado passes by? Due to the difficulties of measuring wind speeds in tornadoes, scientists don't have answers to these questions. However, a collaborative project between researchers at the University of Miami and NCAR has been seeking clues with new, highly detailed computer simulations of tornado wind fields. The simulations can be viewed in a series of animations, created by NCAR scientist George Bryan, that provide a 3D window into the evolving wind fields of idealized tornadoes at different rates of rotation. The "high-swirl animation," shown here, which depicts a powerful tornado with 200-plus mph winds, the purple tubelike structures depict the movements of rapidly rotating vortices. Near-surface winds are represented by colors ranging from light blue (less than 20 meters per second, or 45 mph) to deep red (more than 100 meters per second, or 224 miles per hour). The vortices and winds are contained within a condensation cloud that rises more than 500 meters (1,640 feet) above the surface. Such visualizations can help atmospheric scientists better understand the structures of tornadoes, as well as the shifting location and strength of maximum wind speeds.  Bryan also uses them in presentations to meteorology students. “When you make these 3D visualizations and then animate them, they give you a sense of how the flow evolves and how the turbulence changes,” Bryan said. “These are details you don’t see by just looking at a photograph.” For example, he learned from the visualization that the rotating tubes tilt backward against the flow at higher altitudes. These are the kinds of details that can eventually help scientists better understand these complex storms. The information is also critical for public safety officials and engineers. “If you’re an engineer and designing a building, you want to know details like how much greater is peak wind over average wind in a tornado,” Bryan said. “We’ll get questions from engineers asking about the details of wind gusts in those purple tubes.” Bryan is collaborating on the simulations with Dave Nolan, chair of Miami’s Department of Atmospheric Sciences. To create the animation, Bryan used innovative NCAR software that enables researchers in the atmospheric and related sciences to analyze and interpret results from large computer models. VAPOR (Visualization and Analysis Platform for Ocean, Atmosphere, and Solar Researchers) is an interactive 3D visualization environment for both animations and still-frame images. The open-source software can be downloaded and used on personal computers. VAPOR was developed at NCAR in partnership with the University of California at Davis and Ohio State University. Funding comes from the National Science Foundation and the Korea Institute of Science and Technology Information. Writer/contactDavid Hosansky FunderNational Science Foundation CollaboratorUniversity of Miami


Subscribe to Weather Research