Laserfiche WebLink
QA/QC of ASCE Std ET datasets <br />Automated Replacement of Missing/Bad Data <br />16 <br />Linear Regression Filling of Missing/Bad Data: (y=mx+b) <br />•Uses up to 3 defined replacement stations (in priority) <br />i.e. if day to be replaced is also missing or bad in 1st and <br />2nd stations, looks at 3rd station for that day <br />•Replacement station can also be manually identified <br />•Only compares good to good data for regression statistics <br />(so must know bad data for replacement stations too) <br />•Regressions are currently done on an annual basis due to <br />sensor changes; unless sufficient good data in common is not <br />available in which case it uses full dataset <br />•If absolutely no data available from nearby stations, does a <br />linear interpolation between days with good data. <br />•Regression not used for vapor pressure <br /> <br /> <br /> <br />