BAROMETRIC PRESSURES ON THE GREAT LAKES 131 



frequent periods of 10 days or more on which the observed elevations at a 

 gage are continuously too high or continuously too low to represent the 

 elevation of the mean lake surface. As extreme cases, note the facts for 

 the Cleveland gage in August 1910 and for the Milwaukee gage in July 

 1911. 



At Cleveland (see table No. 20, page 85, table No. 26, page 108, and 

 plates 8 and 9), in August 1910 the correction for barometric effect was 

 + on 25 days out of 31. The maximum correction for barometric effect 

 was +0.24 foot on August 25. On the 5 days August 21-25 the average 

 correction was +0.18 foot. For the whole month of August the corrected 

 elevation was 0.06 foot greater than the observed. 



At Milwaukee (see table No. 21, page 88, table No. 27, page 109, and 

 plates 11 and 12), in July 1911 the correction for barometric effect was + 

 on 28 days out of 31. The maximum correction for barometric effect was 

 +0.30 foot on July 25. On the 7 days July 20-26 the average correction 

 was +0.15 foot. For the whole month of July 1911 the corrected elevation 

 was 0.09 foot greater than the observed. 



The results of this investigation, in the form of corrections for wind 

 effects and barometric effects, and detection of abnormal values by certain 

 criteria, should be applied to regulation of the Great Lakes, (a) directly, to 

 furnish a more accurate knowledge than would otherwise be possible, day by 

 day, of the actual mean elevation of the whole surface of each lake, and there- 

 fore of the total water content of the lake, and, (6) indirectly, by enabling 

 the laws of evaporation and of run-off into the lake to be determined and 

 understood in such wise that a forecast of the total yield of water to each 

 lake for weeks and possibly months in advance may be made with much 

 greater accuracy than would otherwise be possible. 



APPLICATION TO DETERMINATION OF MEAN SEA-LEVEL AND TO 



PRECISE LEVELING. 



Elevations determined by precise leveling are referred to mean sea-level 

 by means of observations taken at tide gages. The mean sea-level as fixed 

 by the observations at a given gage is in error by an amount dependent on 

 the configuration of the shores and the bottom in the surrounding region 

 and upon the prevailing winds. The method and constants now available 

 will enable one to compute the necessary correction to be applied to the 

 observed mean sea-level to eliminate the wind effect and so to obtain the 

 true mean sea-level. Such a necessary correction may be small at certain 

 gages. It is important to prove it to be small in such cases. The 

 corrections are probably large enough at some gages to predominate over 

 the accumulated errors in the precise leveling for hundreds of miles from 

 the gages. 



The corrections for barometric effect should be applied similarly in con- 

 nection with the determination of mean sea-level. 



