Upcoming Winter Season Possibilities Using the Past Summer Temperature and Precipitation Records at Oklahoma City
National Weather Service Forecast Office
This paper will utilize 105 years of temperature and precipitation records from Oklahoma City, Oklahoma (OKC) to examine statistical relationships between summer temperatures and precipitation, and temperatures and precipitation for the following winter season. The goal of this study is to determine whether skill above climatology can be achieved to predict seasonal temperature and precipitation for OKC.
The data for this study were extracted from the Local Climate Data (NCDC, 1890-1996), or LCD, for OKC. Data were extracted from the LCDs for the summer (June, July, August) and winter (December, January, and February) seasons.
The average summer and winter seasonal temperatures for each season from 1891 through 1995 were calculated by averaging the three monthly average temperatures for the summer and winter months. Total precipitation was calculated by summing the three monthly totals for each season. The seasonal average temperature and precipitation were plotted on a distribution chart (Charts 1-4). These plots were subdivided into three equal groups (approximately 33 percent) similar to Namias (1964). The groupings were then categorized as either warmer than normal, normal, or cooler than normal and with respect to precipitation they were labeled as either wetter than normal, normal, or drier than normal.
The extent to which a season was warmer or cooler than normal or wetter or drier than normal was also examined and were labeled as significant events. For example, when examining warmer than normal summers, which consisted of 33 percent of the distribution of all summers, the upper half of that category or 17 percent was defined as significant.
A. Winter temperatures following specific summer temperature regimes.
After warm summers (Fig. 1) there was not a significant correlation to the upcoming winter temperature regime. After normal summers however, their was a distinct trend away from normal temperatures the following winter. The percentage of occurrence of both warm and cool winter temperature regimes increased from the 33 percent that would happen by chance with a slightly higher trend toward a cool winter. After cool summers, 44 percent of the following winters were near normal with respect to temperature.
B. Winter precipitation following specific summer temperature regimes.
Winter precipitation following particular summer temperature regimes (Fig. 2) showed two interesting trends. Again, when looking at the winters following warm summers, just slight changes over climatology occur. However, after normal summers a trend toward dry winters and away from wet winters appeared. Forty-three percent of winters after normal summer temperatures were drier than normal and only 20 percent of winters after normal summers were wetter than normal. After cool summers only 22 percent of the following winters were drier than normal with both normal and wetter than normal winters occurring 39 percent of the time.
C. Winter temperatures following specific summer precipitation regimes.
Upon examination of Figure 3, all three categories indicate only slight variations from climatology in any direction for either warmer, cooler or normal winters after particular summer precipitation regimes. The largest anomaly was noted when examining winters following wetter than normal summers. In this instance, cooler than normal winters occurred only 26 percent of the time.
D. Winter precipitation following specific summer precipitation regimes.
In Figure 4, particular winter precipitation trends did follow certain summer precipitation regimes. After dry summers the following winter tended to be normal with respect to precipitation. The winters following normal summer precipitation regimes were drier than normal 41 percent of the time, while only 27 percent of the following winters were wetter than normal. When wetter than normal summers were examined, 43 percent of the following winters were wetter than normal and only 26 percent of the winters were drier than normal.
E. Winter temperature and precipitation following significantly warmer summers.
Figure 5 indicates that after significantly warm summers, cooler than normal winters occurred only 22 percent of the time. Examination of the winter precipitation after significantly warm summers indicated that 45 percent of the upcoming winters were wetter than normal and only 22 percent of the following winters were normal with respect to precipitation.
F. Winter temperature and precipitation following significantly cool summers.
After significantly cool summers (Fig. 6), winter trends showed an equal likelihood of occurrence with respect to temperature. All three, warmer than normal, normal, and cooler than normal were near 33 percent. This was a substantial change from cooler summers where most of the following winters were near normal with respect to temperature (Fig. 1). Winter seasons after significantly cool summers tended to be wetter than normal with 45 percent of the following winters having this characteristic.
G. Winter temperature and precipitation following significantly dry summers.
Much drier summers had very little impact on the upcoming winter precipitation or temperature regime. Only small deviations from climatology were observed (Fig. 7). Twenty eight percent of the following winter seasons were normal with respect to temperature and the same percentage of winters were wetter than normal after significantly drier summers.
H. Winter temperature and precipitation following significantly wet summers.
The last group examined were winter temperatures and precipitation following significantly wet summers (Fig. 8). As before, temperature trends showed only small deviations from climatology. Precipitation trends did indicate that as summers became increasingly wet, the percentage of following winters being wetter than normal also increased. Following significantly wet summers 50 percent of the following winters were wetter than normal. From Figure 4, after normal precipitation regimes in the summer the following winter was wetter than normal only 31 percent of the time. After wetter than normal summers the following winter was wetter than normal 43 percent of the time.
Several of the groups examined in this study showed skill above climatology in determining the upcoming season's temperature and precipitation regime.
Normal temperatures during the summer tended to favor winters that were both cool and dry. When the summer was cooler than normal, however, the following winter was more likely to be normal with respect to temperature and the likelihood of a dry winter was significantly less. Summer precipitation suggested skill in determining the upcoming season, as well. Following wetter than normal summers, the winters were less likely to be cool; slight increases arose in both the warmer than normal and normal winters. Winters also tended to be wetter than normal following wetter than normal summers. Drier than normal summers were followed by winters with normal precipitation, and normal summer precipitation regimes were followed by drier than normal winters.
Summers in which temperature or precipitation departed significantly from normal also continued some of the trends that first appeared in the previous examinations. The occurrence of cool winters following significantly warm summers continued its downward trend and the trend of wetter than normal winters continued to increase after significantly warm summers. Wetter than normal winters also continued to rise following significantly cooler than normal summers. Winters following significantly wet summers were wetter than normal 50 percent of the time, which is incredible if only 33 percent should happen by chance.
By examining past seasons, trends appear that suggest a correlation between what occurs from season to season. This also presents the possibility that skill can be achieved in determining tendencies of the upcoming season based on the previous season.
NOAA, 1890-1996: Local Climate Data. National Climatic Data Center. [Available from NOAA, NCDC, Federal Building, Asheville, NC 28801-2696.]
Namias, J., 1964: A 5-Year Experiment In The Preparation Of Seasonal Outlooks. Mon. Wea. Rev., 92, 449-464.