BODC HEADED DOCUMENT NO.  41600                         
 Applies to BODC Series Ref. Nos.:   29842,  29854,  29866,  29878,  29891,     
                     29909,  29910,  29922,  29934, 442639, 442640, 442652,     
                    442664, 442676, 442688, 442707, 442719, 442720, 442732,     
    Offshore Sea Floor Pressure Data Screening Carried Out by BODC   1          
    BODC screen both the series header qualifying information and the parameter 
    values in the data cycles themselves.                                       
    Header  information  is  inspected for 1) irregularities  such  as  suspect 
    values;  2) inconsistencies between related information, for example depths 
    of meter and sea bed; times of mooring deployment and for start/end of data 
    series;  length of record or number of data cycles, the cycle interval, the 
    clock  error  and the period over which accrued;  originator's comments  on 
    meter/mooring  performance,  data quality and parameters measured  and  the 
    parameters  actually present in the data cycles.  Documents are written  by 
    BODC highlighting irregularities which cannot be resolved.                  
    Data cycles are inspected via time series plots of parameters.  These plots 
    undergo  intrinsic and extrinsic screening to detect suspect values  within 
    the  data cycles themselves and inconsistencies when compared with adjacent 
    data  sets.  Values suspected of being of non oceanographic origin  may  be 
    tagged with the BODC flag denoting suspect value.                           
    The  following types of irregularity,  each relying on visual detection  in 
    the time series plot, are amongst those which may be flagged as suspect:    
    Spurious data at the start or end of the record.                            
    Obvious spikes of non oceanographic origin.                                 
    A sequence of constant values in consecutive data cycles.                   
    If  a large percentage of the data are affected by  irregularities,  deemed 
    abnormal, then instead of flagging the individual suspect values, a caution 
    may be documented.                                                          
    Inconsistencies  between the characteristics of the data set and  those  of 
    its  neighbours,  as for example between the maximum and minimum values  of 
    the parameters (spikes excluded) are, where necessary, documented.          
    This  intrinsic  and extrinsic screening of the parameter values  seeks  to 
    confirm  the qualifying information and the source laboratory's comments on 
    the series.  In screening and collating information, every care is taken to 
    ensure that errors of BODC making are not introduced.