National Weather Service Forecast Office - Cleveland OH
Harold J. Keeney and Erik S. Pytlak
Correlations are established between occurrences of El Niño/Southern Oscillations and climatic changes in the Ohio Valley and southern Great Lakes region of the United States. During El Niño winters, considered November to March in this study, temperatures usually averaged above normal, while precipitation averaged below normal. Using these correlations, long range weather forecasts, and activities dependent on such forecasts, can be adjusted accordingly.
The phenomenon of sudden warming of the tropical Pacific ocean off the coast of Peru is known as El Niño. This phenomenon occurs roughly every three to seven years, and was once considered a local fluctuation affecting only the coasts of Peru and Ecuador.
Today, the effects of El Niño on world climate are known to be more widespread. Although the El Niño/Southern Oscillation (ENSO) phenomenon is well documented (Trenberth, 1990), the relationship of the ENSO to mid-latitude surface temperatures and precipitation are not covered extensively in the literature (Ropelewski and Halpert, 1986).
The effects of ENSO on world climate are widespread and highly variable. The pattern of ocean temperatures in each event and how the El Niño episode evolves plays a crucial role in weather occurrences across the globe. Some global effects most frequently observed during an ENSO event are: wet conditions during the winter in the southeast United States, drought in eastern Java, southern Peru, and Australia, plus, generally warmer than normal winters in the northwest portions of the United States and Canada (Knox, 1992, Ropelewski and Halpert, 1986).
Ropelewski and Halpert in their 1986 study evaluated North American temperature and precipitation response to ENSO based on a statistical method utilizing the coherence of harmonic vectors. In that study, no clear ENSO response was found in their Mid-Atlantic region that encompasses the area of concern in this study. However, they concluded that the lack of a consistent ENSO response does not rule out the possibility of ENSO-related phenomena. The reason for the inconsistent ENSO response in the Ohio Valley and lower Great Lakes is uncertain. However, weak ENSO events could be masked by other climatic variables such as volcanic eruptions or long term wet and dry cycles (Waters, 1991).
The dynamic response of the atmosphere to an ENSO event has been addressed in many previous studies. Trenberth, in his 1990 study, concluded that the change in diabatic heating can change global general circulation patterns and associated poleward heat flux. A mid-latitude response to ENSO forcing is further supported by theoretical and modeling studies of the circulation (Hoskins and Karoly, 1981, Webster, 1981, Blackmon et al, 1983). Since poleward heat flux is most pronounced during the cool season, winter temperatures and precipitation in North America would most likely be impacted the most by any changes in the circulation pattern.
Therefore, the November-March period of El Niño years in the mid-latitudes would most likely exhibit the most significant ENSO response. Consequently, the focus of this study was primarily on this time period. Indeed, the data set for this period does seem to indicate an ENSO climatic response in the Ohio Valley and southern Great Lakes.
Climatological data covering the last eight ENSO events beginning with 1957 was collected in order to determine whether the climate of the Ohio Valley and southern Great Lakes was impacted by ENSO. Seasonal snowfall, plus temperature and precipitation departures from the mean covering November through March were studied.
Climatological data from 17 first-order FAA reporting stations in Indiana, Kentucky, Ohio, Pennsylvania and West Virginia were collected. The specific stations utilized in the study included: Akron-Canton, OH; Charleston, WV; Cleveland, OH; Cincinnati, OH; Columbus, OH; Dayton, OH; Erie, PA; Evansville, IN; Fort Wayne, IN; Huntington, WV; Indianapolis, IN; Lexington, KY; Louisville, KY; Pittsburgh, PA; South Bend, IN; Toledo, OH; and Youngstown, OH.
These stations were chosen because they all kept climatological records since the 1957 El Niño event. Other first order stations, like Mansfield, OH and Jackson, KY did not meet that criterion.
The data from the seventeen stations was compared to the 30-year averages for 1951-1980, as published by the National Climatic Data Center in Asheville, NC. The anomalies for all seventeen stations were then averaged to obtain a regional anomaly for each El Niño winter.
3. RESULTS AND CONCLUSIONS:
During an ENSO event, winters in the Ohio Valley and lower Great Lakes averaged warmer than the normal. Out of the eight ENSO events studied, the regional average cold season temperature was above normal five times. On the average for the region, temperatures were .45 degrees Fahrenheit above normal.
Winter precipitation showed a deficit in seven out of the eight ENSO events studied. The average precipitation deficit in the region studied was 1.82 inches. This led to below normal snowfall at many stations. This was particularly pronounced at stations removed from lake effect snow regimes.
Lake effect snow and orographic effects may mask the ENSO response in some years at stations in the vicinity of the Great Lakes and Appalachian Mountains. Indeed, no correlation between years with below normal snowfall and El Niño years can be obtained from stations like South Bend, IN, Cleveland, OH, Youngstown, OH, and Charleston, WV. All of these stations either experience lake effect snowfall to some degree, or are subject to some orographic effects. It appears through this study, that these and possibly other mesoscale climatological effects are not appreciably influenced by El Niño events.
Based on the data collected, strong El Niño events are most likely to affect the climate of the Ohio Valley and lower Great Lakes. Weak El Niñoes, like the 1969-70 and 1976-1977 events, may cause other anomalous climatic affects like colder than usual winters. Unfortunately, the data sample used in this study was too small to make such a conclusion.
However, the frequency of climatic response to strong and moderate ENSO events is high in the Ohio Valley and southern Great Lakes. During a moderate to strong ENSO, winters in the Ohio Valley and lower Great Lakes are likely to be warmer and drier than normal. This in turn results in below normal snowfall across the region, aside from lake effect areas.
A more complete study would involve investigating warm season precipitation and temperatures for the same region. Also, it might be worth looking at annual temperature and precipitation anomalies.
The frequency of an ENSO response is apparently high enough so that ENSO considerations should be incorporated into long range winter outlooks for the Ohio Valley and lower Great Lakes, especially if a strong El Niño is in progress. This information is available via the Climate Analysis Center at the National Meteorological Center.
More accurate winter outlooks would benefit many weather sensitive industries. For instance, highway departments could budget less money for snow removal and overtime during ENSO years. Conversely, ski resorts would have to budget more money for snow making and be prepared for lower profits. Meanwhile, natural gas and oil distributors could reduce inventories.
Below normal winter precipitation and snowfall could adversely impact some agricultural operations. Specifically, winter wheat yields the following summer could be reduced due to deficient moisture and possible cold damage because of the absence of an insulating snow cover. In fact, the incipient stages of some Ohio Valley and lower Great Lakes droughts may be initiated during a dry El Niño winter.
Edited by Herman Washington