Laserfiche WebLink
<br />l <br /> <br />tempera ture may be selected at -2OG or -2lC. Add it ional det. Us j n r ""ard " t h" <br />Hamilton Branch curve may be found in Section 2.4.2. <br />1.2.6 Data Sources. All computer-ccIDpatible data for use in this study has <br />been obtained from tbe National Climatic Center (NCC), Asheville, North Carolina. <br />All such data was ohtained on seven-track magnetic tape, 556 bpi, even parity, <br />blocked ten card images per physical record. <br />The data used for the determination of precipitation episodes was found in <br />the publication series Hourly Precipitation Data. The time lags required to relate <br />precipitation episodes with appropriate rawinsonde observations were determined <br />by analysis based on the Synoptic Weather Maps, Daily Series (Part 1, Northern Hemi- <br />sphere Sea-Level and 50G-Millibar Charts) for the periods they were available and <br />based on the Daily Weather Maps, Weekly Series for the remainder of the study period. <br />All three of the Environmental Data Service publications cited above were available <br />in the NAWC technical library. <br />1. 2. 7 Da E'- P r DC es s iEA . The c mapu terized da ta processing r eq ui red for this pro- <br />ject can best be considered in tWO parts. The first part include. editing the input <br />data, mergiog the various data sources, and deriving the parameters to be used in <br />the analysis. The product o[ this portioo of the processing is a data file which <br />includes, in a convenient form, all of the parameters required [or the analysis. <br />This file is termed the master file. The second part of the data processing encom- <br />passes all of the various summarizations performed on the master file. <br />1.2.7.1 product.ion of master fi1(!s. This portion of the processing is by <br />far the most expensive and time consuming. This is primarilY a result of the very <br />large volume of data that must be dealt with. The objective o[ this effort is to <br />distill the voluminous input data down to a concise, manageable file whid, would then <br />allow the analysis effort to proceed unencumbered by extraneous data or data [ormat- <br />ting problems. <br />Figure 1. 2-10 shows a simplified block diagram o[ the process used to obtain <br />the roaster file for a given massif. The precipitation episode specifications and <br />the time lags required to associate precipitation with appropriate rawinsonde ob- <br />servations are punched on data processing cards and used to select the soundings <br />which are needed for the remainder of the ann1ysis. These selected soundings nre <br />then input to the sounding analysis program which calculates all the parameters to <br />be used later in the analysis. Finally, each precipitation observation is associat- <br />ed, if possible, with a sounding representative of the air over the precipitation <br />station as specified by the rawinsondp time lags. <br /> <br />I <br /> <br />1-29 <br /> <br />I <br />