Reruns for Jan, Jul 1953

On the left are the corrected minus uncorrected monthly mean surface pressure fields, while on the right are the monthly mean MSLP fields from the corrected run, with the areas of <1000mb shaded in green. The differences on the left have some correspondence with the shaded <1000 mb areas, since the error corresponded to incorrectly coding values of sfc press and mslp when the values became <1000mb.

The error is absent in the January Aleutian Low but is present end of the Icelandic low; this disparity arises from Aleutian low being defined by two unaffected sources US islands and ships, while the miscoded surface pressure obs strongly influence the circulation over the Barents Sea.

The difference in July over the Mid-East and over East Asia are due to the constancy of the Asian Monsoon heat lows. Providing the availble obs corrects for model bias.