33 / 28
      34 / 29
      40 / 26

      Preview of the Coldest October: 16th in 2002

      The following winter was a cold one as a eastern North American trough dominated.

      A record low of 22 was recorded at the National Weather Service (NWS) near Negaunee on October 16, 2002. This low occurred near the beginning of the run-up to the coldest October on record. The month started warm with a high of 72 on the first. However, like this month is expected to do, the pattern flipped to cold. Consistently below average temperatures then persisted through the end of the month as a deep upper-air trough formed from northern Canada into the Upper Midwest (Image 1 above). The trough kept â??reloadingâ?? as a big ridge planted itself near the West Coast. A number disturbances dropped southeastward through the trough into Upper Michigan. The main one on October 21-22, brought heavy precipitation to a good share of the U.P. The air was cold enough for snow. Even in the south-central U.P. Iron Mountain received a 24-hour record 8 inches between the 21st and 22nd. To the north at the NWS, Lake Superior enhancement brought over a foot of snow on the same days.

      October 2002 ended 5.3 degrees below average. The 37.9 degree mean edged out October 1988 for the coldest tenth month at the NWS. November continued the cold theme coming in nearly four degrees below average. While December started very coldâ??the first five days came in a bitter 15.8 degrees below averageâ??the pattern flipped to warmer during the second week. The warmth then held on through Christmas well into January. However, the pattern again flipped to colder and the winter as a whole was harsh as the eastern North America trough dominated (Image 2).

      The first portion of this month is running much above average. Three days were 20 degrees or more above average. Through yesterday, the mean temperature was 6.6 degrees above average. The pattern evolving now will cut back significantly on October 2013â??s sky-high anomaly. A huge upper-air trough is forecast by all models to develop over central North America. This will mean a much colder pattern for Upper Michigan. One valuable long-range forecasting technique is the analog method. A forecaster looks at the predicted pattern and compares it to patterns in the past. The Climate Prediction Center (CPC), a branch of NOAA, puts out a computerized analog package for 8 and 11 days into the future (Image 3 & 4). Todayâ??s 8-day package based on the GFS Ensemble (a synthesis of a number of model runs each with different initial conditions) shows a deep negative upper-air anomaly (blue) just northeast of the U.P. To the right is the top analogs that match this forecast pattern most closely. The top analog for both days is October 20, 1976.

      Looking back to that time, Upper Michigan was experiencing an exceptionally cold pattern. In Marquette, the seven following days had three days with highs of 40 into the low 40s with the other four days only in the mid to upper 30s. There was measurable snow on four of those days for a total accumulation of 10.7 inchesâ??a rare event in October at lake-level.

      Of course, an analog will point us in the direction the pattern may be heading; it is not to be used to forecast the same weather that occurred 37 years ago. At the same time, however, this is a sign that we are likely to see some early winter weather at times over the next couple of weeks.