Midwestern Corn Yield and Weather in Relation to Extremes of the Southern Oscillation

Weather and crop yields in the Midwest have exhibited wide fluctuations during the past 10 yr. Sea surface temperatures (SST, El Niño and its counterpart, La Niña) have been related to or blamed for these weather abnormalities via teleconnections. Yield, weather, and El Niño related (southern oscillation index — SO) data beginning in 1900 were assembled to determine if significant relationships could be ascertained. Midwestern weather and corn (Zea mays L.) yield data were coded relative to the SO. The SO groupings were <−8.0 (El Niño like, low phase), <0.8 (La Niña like, high phase), and in between. Yields were grouped >10% or <10% relative to expected, or in between. Corn yields exhibited wide variation when the SO was in between, indicating that weather factors other than SO influenced corn yield during those oceanic conditions. However, when summer SO was in the high phase (low phase), there was a statistical tendency for corn yields to be lower (higher) than expected, respectively for all Corn Belt states studied, except Missouri. The low (high) phase of the SO is generally related to El Niño (La Niña). During the low (high) phase of the SO, much of the Corn Belt received more (less) rainfall in July, August, and September. At the same time, high temperatures—or heat stress—were generally lower (higher) during the low (high) phase of the SO. Both high (but not excessive) precipitation and lower temperatures are associated with good corn yield. If El Niño forecasts improve as is expected, Midwestern corn yield forecasts should similarly improve. Research QuestionIn recent years, much has been said about temperatures in the Pacific Ocean and weather in the USA. The recent warm ocean event (El Niño) started in 1991 and continued until early 1995. This event was unusually long when compared with similar events during this century. During this same time, great yield variations have been experienced in the Midwest. We examined the statistical relationships between this ocean phenomenon and both Midwestern corn yields and weather. Literature SummaryIt has been shown that sea surface temperature (SST) anomalies can be associated with both positive and negative weather anomalies throughout the world. This includes both temperature and precipitation variations. The oceanic anomalies have been related to air pressure differences between the central Pacific Ocean (Tahiti) and Australia (Darwin). This difference is expressed as the Southern Oscillation (SO). The SO influences both winds and ocean currents, and they, in turn, produce equatorial SST anomalies in the Pacific. In fact, these SST anomalies can then influence pressure and winds in this area by changing the fluxes of heat and moisture to the atmosphere. The latter affects solar radiation and other atmospheric variables by producing clouds and possible precipitation. El Niño and La Niña are termed warm and cold events, respectively. The El Niño is combined with the low (negative) phase of the SO and termed ENSO in meteorological literature. La Niña generally corresponds to the positive phase of the SO. Study DescriptionWeather and corn yield data for Iowa, Illinois, Indiana, Missouri, and Ohio and SO data were assembled for 1900 to 1994. The weather data included monthly maximum and minimum air temperatures, precipitation, aridity, and heat stress. Supplemental monthly average temperature and precipitation data were compiled over the north central region of the USA from 1890 to 1991. Corn yield data for each state were fitted with a nonlinear regression equation to calculate an expected yield value for each year. This was necessary to account for complex technology trends in corn yield over this period and put year-to-year yield values on a relative basis. The weather and corn yield data were grouped according to the occurrence of monthly SO index values greater than 0.8, less than −0.8, or in between. Our intent was to examine these relationships when rather extreme oceanic and atmospheric conditions were prominent in the South Pacific. Mean values of grouped data were tested for significant differences from long-term values by using the t-test. Applied QuestionsDoes the SO have an impact on Midwestern crops? Analysis of long-term Iowa corn yield data shows good stability with 57% of the years within ±10% of expected. Twenty-five and 18% were above and below the +10 and −10% levels, respectively. When the summer SO levels were less than −0.8, 86% and 32% of the yields exceeded the −10% and + 10% yield levels, respectively. Thus, severe reductions in yield tend to be avoided under those circumstances. The 14% of the years in this SO grouping with substantial yield reductions occurred under unusual circumstances. Excessive wetness occurred in 1993, early frost damage was evident in 1951, and 1977 exhibited extreme midseason weather anomaly reversals during critical reproductive stages. When the summer SO values exceeded 0.8, 93% of the years were + 10% or lower. In fact, 40% of these years were below the −10% yield level. Only 1 yr exceeded the 10% yield plateau. The statistical impact of SO on the yields observed in other Midwestern states was comparable to the Iowa response with Missouri showing some exceptions. Thus, SO is not the sole yield-determining weather-forcing function in Iowa or the Corn Belt. There is, however, a statistical tendency for above normal yields with negative summer SO values and below normal yields with positive summer SO values. Does the SO influence Midwestern weather? Long distance weather relationships like this across the earth are termed teleconnections. We found that weather conditions in Iowa were more favorable for corn production when the summer SO values were less than −0.8, because, on average in Iowa, more rain is received and air temperatures are less extreme. However, it should be stressed that these relationships are purely statistical. There is no clearly identified physical relationship between the SO and Midwestern weather, although progress is being made. The temperature and precipitation response was, however, variable over the Corn Belt. In Iowa, heat stress and aridity were reduced for less than −0.8 SO summers when compared with years with summer SO values exceeding 0.8. These two variables were not computed for other states in the Midwest.