Factors Affecting Ship and Buoy Data Quality: A Data Assimilation Perspective

Abstract

Ship and buoy reports of wind, air pressure, temperature, humidity, and sea temperature for 2007 and 2008 have been compared with values from the operational Met Office global numerical weather prediction (NWP) system. Ship reports have been categorized by vessel type, recruiting country, and manual or automatic reporting. Most estimated ship winds (except Dutch ones) are too strong and are better treated as measured winds. After height adjustment, the average ship wind speeds are reasonably consistent with buoy wind speeds, but the global model winds are 6%?12% weaker. Extra care is needed when comparing/assimilating values close to the coast. Ship air temperatures are too warm during the afternoon due to solar heating of the ship. Adjustment of the pressure to sea level is a problem for some larger ships. Passenger ship reports are relatively poor quality for several variables and their winds are rather strong. Automated ship reports and those from research and coast guard vessels tend to be of good quality but some of the winds are slightly too weak. In most respects buoy data have the best quality. A number of improvements to the Met Office observation processing system are being made, notably tightening of the quality checks and better height adjustment of winds.