The Automated Precipitation Correction Program

The Automated Precipitation Correction Program (APCP) The Automated Precipitation Correction Program (APCP), written in Visual Basin 6.0, is developed to remove mechanical noise signals from the high frequency (15 or higher) precipitation data collected using the weighing-recording universal precipitation gauges APCP is designed to preprocess one year of precipitation data at a time Userdefined parameters for the APCP include BucketDecanting Limit, BucketRecharge Limit, Noise Limit and Out of Range Value Identifier Weighing-recording gauges are most commonly used to record high frequency precipitation data in environments which receive substantial precipitation as snow The Single Gauge Version of the APCP can be used to preprocess precipitation data from the sites having only one precipitation gauge Input file for the Single Gauge Version of the APCP should contain time series of precipitation data with Julian Day, Time (2 digit hours and digit minutes; for example 1330 represent 1:30 PM) and measured precipitation (mm) in space or comma separated columns Measurements of snowfall by the precipitation gauges is severely affected by the undercatch related to systematic inefficiencies The Dual Gauge (system of precipitation measurement was developed to better estimate actual ground precipitation and used at the Reynolds Creek Experimental Watershed, Dry Creek Experimental Watershed and many other locations The dual gauge installation consists of two universal weighing recording precipitation gauges, one unshielded and the other shielded with Alter-type shield, located in close proximity The Dual Gauge Version of the APCP is developed to remove mechanical noise signals from the precipitation data collected using dual gauge installations Input file for the Dual Gauge Version of the APCP should contain time series of precipitation data with Julian Day, Time (2 digit hours and digit minutes), shielded gauge catch (mm) and unshielded gauge catch (mm) in space or comma separated columns (Input as rfp, output rfc) precipitation using the Dual gauge algorithm The input file for the Wind Correction program should contain hourly time series of dual gauge precipitation data after removal of mechanical errors with Julian Day, Time (2 digit hours and digit minutes), shielded and unshielded gauge precipitation (mm) in space or comma separated columns This input file can be generated by selecting Fixed Interval (min) option and using 60 in adjoining text box of the Dual Gauge Version of the APCP Readme: Input files contain Day of Year, Time of day (two digit hr and two digit min; for example 1030 is 10:30 AM), Shielded and Unshielded gage catch in respective columns The input file (176.2006.rfp) was corrected for Mechanical errors using default values in APCP The corrected file is named as '176.2006.rfc' The wind correction is done using a wind correction program that implements one of the following methods: (1) Dual Gauge (Hamon 1971, 1973) (2) DFIR Intercomparison (WMO) for shielded Gauge (3) DFIR Intercomparison (WMO) for Unshielded Gauge The 'Dual Gauge' wind correction program can take hourly output file generated using APCP (with DayofYear, Time (HrMn), Cumulatvie Shielded, and Cumulative Unshielded data in respective columns) as an input file Here output file using 'Dual Gauge' method is named as '176.2006.DG.rfc' The coefficient B = 1.8 was used to compute wind corrected values The output file generated using 'Dual Gauge' technique contain: DayofYear, Time (HrMn), Shielded, Unshielded and WindCorrected data in respective columns 'DFIR intercomparison for shielded gauge' techniqe requires input file with DOY, Time, Shileded Gauge Catch, Wind Speed (m/s), and Temperature (or Dew Point Temperature) (C) in respective columes The output file contains: DOY, Time, Shileded Gauge Catch, Wind Speed (m/s), Temperature (or Dew Point Temperature) (C), Wind corrected ppt, and Phase in respective columens The calculations are done using the equations presented in WMO 1998 report (Goodison et al 1998) and Yang et al (1998) but restricting maximum daily wind speed to 6.5 m/s to avoid unreasonable wind corrected valuse 'DFIR intercomparison for unshielded gauge' technique requires input file, and generates output file similar to the 'DFIR intercomparison for shielded gauge' technique, but using Unshileded gauge data (in 3rd column of input file) instead of shielded gauge data Notes by Pam Aishlin regarding data: G:\Boisefront\hypsometric\Filtered Precip Data\corrected precip99_06 The precipitation filter created by Anurag Nayak of Utah State University, 2005, in collaboration with USDA ARS in Boise, has two versions The early version requires input of two complete calendar years and outputs a complete water year The second version works with any single input file and outputs the same time series as filtered data The first version seems to output files with a rfp extension, while the second version outputs with a rfc extension, using rfp as required input extension Some output files have been furthered processed to csv or xls format The rfp or rfc files are space or comma delimited, not tab delimited The filter does not work with tab delimited Output files to year 2004 were produced by Anurag 2004-2006, 2007 by Pam The output files produced by Anurag have not been assessed in detail for error See also source of output files, executables, and data in folders: G:\Boisefront\hypsometric\Filtered Precip Data\pcpfilter_multi year filter_process\Filter_data19982004\boisefront\corrected precip And G:\Boisefront\hypsometric\Filtered Precip Data\pcpfilter_partial year filter_process\2004_present_ data and results August 2008 update It appears, that, unkowingly to the original users, this filter version with instructions for changing date from calendar form to the form of julian day, hhmm, as needed for the filter, produces an error in datetime The dates in this file, and probably others provided by Anurag early in the filter creating process, are in error, usually by 12 hours offset For instance, in the 20042005 output file, the actual date from which data begins is Sep 29, 2004, rather than the supposed Oct 1, 2004, which will be day 275 USE THE NEW FILTER WHICH HAS IMPROVED COLUMN CONFIGURATION AND OUTPUT, rather than these "final filtered output files" For these original output files: The columns of the file contain Day of the year, Day of the water year, time of day, original shielded data, original unshielded data, shielded and unshielded data after bucket dump correction and last two columns contain finally corrected shielded and unshieded precipitation data In assessing 20042005: In this original filter version, the final cum sh and unsh are very similar in value, and are also similar to the corr unsh, THEREFORE, these values are likely low USE OF NEW FILTER IS SUGGESTED  ... precipitation using the Dual gauge algorithm The input file for the Wind Correction program should contain hourly time series of dual gauge precipitation data after removal... errors using default values in APCP The corrected file is named as '176.2006.rfc' The wind correction is done using a wind correction program that implements one of the following methods: (1) Dual... 2004, rather than the supposed Oct 1, 2004, which will be day 275 USE THE NEW FILTER WHICH HAS IMPROVED COLUMN CONFIGURATION AND OUTPUT, rather than these "final filtered output files" For these