REMARKS:
The 1823 Ft. Snelling climatological record consists of fixed time temperature readings taken daily at or about 0700, 1400 and 2100 hours local solar time; single daily entries indicating the prevailing direction of the wind and the general condition of the sky; intermittent records of snowfall and/or snow cover; episodic records of phenological, hydrological, astronomical and/or other natural events (windstorms, prairie fires, etc.); descriptive entries indicating the general duration (and, in some instances, the intensity) of precipitation; precipitation type ; and special atmospheric phenomena (fog, smoke, etc.). So far as can be determined, all 1823 observations were taken within the permanent fort enclosure (on the bluffs overlooking the confluence of the Mississippi and Minnesota [then the St. Peters] rivers) .
Analysis of the relationship between 0700, 1400 and 2100 readings suggests that the station thermometer was almost certainly exposed to direct morning and/or afternoon sunlight during the warm months of 1823. The most obvious instance of sun contamination is indicated by the June 1823 record: according to a visiting explorer the station thermometer was exposed to the direct rays of the afternoon sun, thus providing a probable explanation for the uncharacteristically high readings recorded during that month (and probably during the month of July as well). The 1823 record also provides evidence of early morning sun contamination, indicating 0700 readings that exceeded or equalled 2100 readings on many days in late May and in June, July and August (thereby reversing or nearly reversing normal summertime diurnal patterns). All of which strongly suggests that 1823's warm season average temperatures are significantly distorted: the original record indicates, for example, that June 1823 temperatures averaged about 73 F (as adjusted by Fisk, cited below), a value which, when further adjusted for the probable effects of sun contamination , is reduced approximately 5 F to an average of about 68F! Similarly, the original monthly average for July (75 F) 1823, when adjusted for the distorting effects of sun "pollution", is reduced approximately 2 F to an average of about 73 F.
The foregoing 1823 temperature record includes both unadjusted (UNADJ) and adjusted (ADJ) mean temperatures. Unadjusted values are the averages of fixed time readings taken daily at 0700, 1400 and 2100 hours . Adjusted averages are from Charles J. Fisk's 1984 "Reconstruction of Daily 1820-1872 Minneapolis-St. Paul Temperature Observations". These values are based on statistically derived estimates of the daily maxima and minima that would have been recorded had the Ft. Snelling station been equipped with self-registering thermometers read and re-set at midnight . The foregoing 1823 record also includes both the monthly and annual extreme temperatures (highest daily minimum, lowest minimum, etc.) estimated by Fisk and the monthly extremes actually recorded/observed (OBSRV) by fort observers. All 1823 temperature distributions (e.g. days 90 F or higher, 32 F or lower, etc.) are based on Fisk's estimates of daily maxima and minima.
Although no quantitative precipitation records were kept at Ft. Snelling during 1823, extant records do, as noted, contain entries indicating precipitation type and, in some instances precipitation intensity (e.g. light snow, heavy rain, etc.). These records are the basis of the foregoing 1823 precipitation frequency record (e.g. number of days with snow, etc.) . Prevailing wind values, similarly, are based on entries indicating the predominate direction of the wind on each day of the month. Prevailing monthly winds indicate the direction most frequently observed/recorded during any given month.
Windy, sunny (twenty clear days noted) but snowy January: snow on ten days with five inches recorded on 4 January and six inches recorded on 20-21 January. Reading of -28 F at 0700 hours on 5 January. Moderate mid and late January temperatures followed by a very cold February (5 F). Temperature of -26 F recorded at 0700 hours on 14 February; temperature of -22 F noted at dawn on 27 February. Windy, sunny (twenty two "clear" days recorded) March. Five inch snowfall noted on 10 March. Windy and probably moderately dry April: temperatures in the 70's F at mid-month. Dry, windy early May: rainy conditions during mid-May. Warm late May. Fourteen "cloudy" days recorded during May. Warm, dry June and July: 100 F recorded at 1400 hours on 13 June; 98 F at 1400 hours on 15 June; 97 F at 1400 hours on 16 June; 96 F at 1400 hours on 4 July . Twenty two "clear" days during June, twenty one during July. Southwest winds prevailing, 12-17 June. Southerly winds on twenty days during July. Moderately wet, cloudy (eighteen "cloudy" days) August. Several heavy rainfalls noted: "fresh" winds on many days during August. Warm early September followed by heavy frosts, 25-29 September. Minimum temperatures probably in the 20's F, 25-28 September. Light frosts noted on 7, 12 September. Heavy rains on 1, 2, 9 September; dry late September. Twenty seven "clear" days during September. Warm early October: temperatures in the 80's F on 1, 2 October. "Fresh" winds noted on many days during October. Twenty four "clear" days with generally dry conditions during October. 75 F at 1400 hours on 27 October followed by cold and two inches of snow at the end of the month. Sunny (twenty "clear" days) November. 60 F recorded at 1400 hours on 22 November. Cold, dry December ("light" snow on 8 December, the month's only recorded snowfall). Very cold, 23-25 December (-20 F at 2100 hours on 24 December): "fresh" winds on 23 December. Late December warm-up: 34 F at 0700 hours on 30 December.