58GS20150410 QC and workup notes Are Olsen June 2017 This document has three sections, one for Bottle, one for surface and one for CTD data ******************************************************************************************** BOTTLE ******************************************************************************************** Started with /MergedFromRike/Current/data_58GS20150410_current.txt __________________________ EDITS and NOTES BEFORE QC __________________________ theta, aou and sigma0-3 columns deleted Changed format and column headers to WHP EXCHANGE. Station 193 renamed to 192 Cast 2 BTLNBR_FLAG_W for BTLNBR 1 set to 3 by Friederike. Reset to 2, change later. CTDSAL_FLAG_W came with all flags 2. SALNTY_FLAG_W came with all flags 2. CTDOXY_FLAG_W came with all flags 2. OXYGEN missing values set from NaN to -999, flag from 4 to 9 SILCAT below detection level set to NaN with flag to 2 by Rike. Changes to 0 with flag 2 (only two instances, STATION 206 BTL 4 & 5). NITRAT came with all flags 2. NITRIT below detection level set to NaN with flag to 5 by Rike. Changed value to 0 and flag to 2, make a note in header (many instances). PHSPHT came with all flags 2. CFC-12 and SF& came with QC notes from Emil implemented as flags. DIC and TALK 181/1 TCARBON changed from 2161 to 2164.5 The Carbon chemistry had three replicates, all three used in Rikes file, but final replicate clearly too low so I removed that one and calculated new average. Otherwise all flagged 2 or 6 Note, the readme file for TCARBON and ALKALI data notes several values should be set to NaN. But all TCARBON values appears to have been kept in file, and all flagged 2. Checked the carbon 'master file', these had been measured twice and bad duplicate removed. Only two ALKALI values are NaN, these were not present in carbon chemistry master file, so obviously not measured DEL13C all flags kept. All remaining NaNs changed to -999. Set all missing data flags to 9. _______________ QUALITY CONTROL _______________ Flags: Station 175, 176, 178 Bottle 1, bottle flags set to 3, as leak suspected. Flags for all Niskin sampled variables set to 3. (Note same bottle was used in position 1 for station 175-181, but large difference in CTDOXY-OXY only visible for stats 175, 176, 178) 176 btl 5 Oxygen high vs CTDOXY and CTDPRS, flag 3. 177 btl 9 Oxygen high vs CTDOXY and CTDPRS, flag 3. 179 btl 5 Oxygen high vs CTDOXY and CTDPRS, flag 3. 180 btl 9, ALK high vs SALNTY and CTDPRS, flag 3. 185 btls 2,4,7 PO4 low vs NO3 and CTDPRS, flag 3 189 btl 4 Oxygen high vs CTDOXY and CTDPRS, flag 3. 190 btls 5 & 6 Oxygen likely swapped during work up, swapped. 192, Cast 1 btl 4 Oxygen high vs CTDOXY and CTDPRS, flag 3. 194 btl 4 & 9, ALK very high vs SALNTY and CTDPRS, flag 4. 198 btl 4 Oxygen high vs CTDOXY and CTDPRS, flag 3. 202 btl 3 Oxygen high vs CTDOXY and CTDPRS, flag 3. btl 4, ALK very low vs SALNTY and CTDPRS, flag 3. saved as 58GS20150410_hy1.csv ******************************* At Princeton late June through early August 2018: A very few additional values flagged 3 or 4 via normal QC procedures by Key Cast time extracted from CTD files and merged into bottle data file Revised CTDSAL (except for station 209) and SALNTY values merged into bottle data file Printed and all pertinent files submitted to CCHDO on 8/9/18 ******************************************************************************************** SURFACE ******************************************************************************************** Started with MergedFromRike/Current/data/data_58GS20150410surf.csv added 13C data, from Surface13C_Marie. All data with high standard deviation among replicate analysis deleted All NaN replaced by -999 Changed headers, added units saved as 58GS20150410_surf.csv ******************************************************************************************** CTD ******************************************************************************************** Started with MergedFromRike/Current/data/data_58GS20150410ctd_current.csv did nothing saved as 58GS20150410_ctd.csv