Hakuho Maru Quality Control Report Christopher Harvey Tropical Ocean Global Atmosphere (TOGA) Coupled Ocean Atmosphere Response Experiment (COARE) Surface Meteorological Data Processing Center Center for Ocean Atmospheric Prediction Studies The Florida State University July 28, 1995 Report COAREMET 95-7 Version 1.0 Introduction: The data referenced in this report were collected from the research vessel Hakuho Maru's (call sign JDSS) Marine Meteorological Observation System (data acquired from the Ocean Research Institute - University of Tokyo) for TOGA COARE during the Intensive Observing Period (IOP). The data underwent postcalibration and correction before arriving at F.S.U. and information on these corrections are found in the appendix. All data were converted to a standard format and were then preprocessed using an automated data quality checking program. The data were then visually inspected by a data quality analyst who reviewed, modified, and added appropriate quality control (QC) flags to the data. This report summarizes the flags for the Hakuho Maru data sets, including both flags added by the preprocessor and the analyst. Statistical Information: The data from the Hakuho Maru during the IOP were expected to include averaged observations every fifteen minutes from a cruise which began on November 10, 1992 and lasted for 21 days. Time (TIME), latitude (LAT), longitude (LON), platform heading (PL_HD), platform speed (PL_SPD), wind direction (DIR), wind speed (SPD), atmospheric pressure (P), temperature (T), specific humidity (Q), sea surface temperature (TS), and rain rate (RRATE) were quality controlled. A total of 24,192 values were reviewed and checked and 54 flags were added resulting in 0.2232% of the data being flagged. Table 1 details the distribution of flags and includes the percentages flagged for each variable sorted by flag type. Table 1: Frequency of Flags Assigned for Each Variable and Flag Type Not Within > 4 S.D. From Percent of Realistic Climatological Total Number Variable Records Variable Range Mean of Flags Flagged TIME 0 0 LAT 0 LON 0 PL_HD 0 0 PL_SPD 49 49 2.4306 DIR 0 0 SPD 5 5 0.2480 P 0 0 T 0 0 Q 0 0 TS 0 0 RRATE 0 0 Total 49 5 54 Percent of Total Data Flagged 0.2025 0.0207 0.2232 Summary: As the statistical information indicates, the data from the Hakuho Maru were in very good condition. The only major problem occurred in the platform speed. The 49 values that were not within the realistic range were flagged from approximately -0.1 m/s to approximately -1.0 m/s. The reason for these negative speed values remains unclear. Final Note: The data from the Hakuho Maru are in very good condition and the analyst does not foresee any problems in using the data. Reference: da Silva, A.M., C.C. Young and S. Levitus, 1994: Atlas of Surface Marine Data 1994, Volume 1: Algorithms and procedures. NOAA Atlas Series. In preparation. Appendix: The following information, extracted from the "README" file for the data, details the corrections made to the surface meteorological data from the research vessel, Hakuho Maru. These data were calculated from the original surface meteorological data of the Hakuho Maru in every minute. Data were averaged over fifteen minutes and corrections were applied as follows. The data obtained as routine meteorological data on board the Hakuho Maru are valuable obtaining radiation and energy fluxes. However, by cross checking them with another data set, it was found that they include some measurements errors. The sea surface temperature (SST) was measured at the depth of four meters at the intake valve for the engine water. It was found that SST is almost constant and compatible with bucket measurements when the ship is running. However, when the vessel stops, SST suddenly increases about one degree and shows unstable behavior. It becomes stable again when movement resumes. At the station on the equator, the data collectors repeated upper air sounding, eddy flux measurement, ocean mixing measurement, and CTD cast every three hours. The ship steamed up against the wind to reduce the effects of shop body during the eddy flux measurements. The vessel stayed there during ocean measurements. It was found that the peculiar SST behavior is consistent with ship speed. The physical explanation of this behavior is not certain but it suggests that the ship body could be a heat source for SST even at night. During the periods when the ship speed was less than three knots, SST was deleted and interpolated before and after the period. In order to avoid discontinuity or a spiky signal, a low pass filtering was applied to the corrected data set as SST is considered to have a slow time response. The air temperature was measured on the upper deck of the ship which is rather inside the body. Another air temperature sensor was in the eddy correlation system at the top of the foremast. The deck thermometer data were obtained continuously but the mast sensor was operated every three hours during eddy flux measurements. They were both aspirated and the data were compared when both data were available. It was found that the daytime deck temperatures are about 0.5 degrees height than the mast data and they are almost compatible at night. During the eddy flux measurements, the ship steams up against the wind and the foremast is the most favorable place because it is not affected by the body of the ship. In the MONEX cruise, Kondo and Sato (1979) reported that the air temperature difference between upper deck and the head of the vessel was 0.4 degrees during daytime and -0.1 degrees at night. According the regression analysis between the temperature difference (DT) and solar radiation (DSR), the authors found the following formula. DT = 0.00108(DSR) - 0.1. The air temperature data were corrected based on this equation. The humidity was measured as dew point temperature on the upper deck. It was compared with foremast data, which measures relative humidity with air temperature. There were no systematic differences and no corrections were applied. The wind speed and direction measured on the pole of the upper deck were compared with foremast data measured with sonic anemometer with the ship corrections. They were also consistent and wind data were not corrected.