Observing System Simulation Experiments (OSSE) on the JAMSTEC buoy system
Akimasa Sumi
(Center for Climate System Research, University of Tokyo)
Masahide Kimoto
(Center for Climate System Research, University of Tokyo)
Nobuo Masumoto
(Department of Earth and Planetary Sciences,
University of Tokyo)
1.Introduction
In order to investigate a variability of climate system, a monitoring system for climate system is indispensable. For the tropical atmosphere-ocean system, TOGA TAO buoy network has been designed and implemented due tithe multi-national collaborations (Hayens et al.,1989). However, it costs much to maintain the buoy network system., and we need further enhanced international collaboration.
Based on these situations, JAMSTEC decided to develope and maintain the buoy network for observing atmosphere and upper ocean, which will be supplement for the existing TAO array.
However, as the costs are extremely large, the design should be economical and optimal for the missions. For this purpose, we should evaluate its performance and determine the distribution and sensors of the buoy system.
For this purpose, OSSE(Observing System Simulation Experiments) has been conducted when FGGE was designed. Therefore, the similar experiment is conducted and results will be reported in this report.
2.Experimental procedures
In the OSSE, the truth and the observation are given by the high Resolution model. In our experiments, OGCM of the Department of Earth and Planetary Sciences was used, which is the 21 level sigma-coordinate model with 1/3 degree horizontal resolution in latitude and longitude except for the open ocean(Masumoto and Yamagata,1993). This model is integrated from 1980 to 1988 by using monthly averaged ECMWF wind, temperature and humidity data. This model results are used as " observations" and "truth" vales.
Ocean Data Assimilation system (ODAS) applied in this study has been developed by Kimoto et al.(1996) at JMA. Its horizontal resolution is 2.5 degree in latitude and 0.5 degree in longitude between 10N and l0S and 2.0 degree beyond 20N or 20S.
Following experiments are conducted and their results are compared (see,Fig. 1).
(1) exp00:ODAS model is driven by climatological wind stress.
(2) exp01:ODAS model is driven by wind data on the dense surface network,
(3) exp02:Surface temperature data and wind data are assimilated over the entire tropical region,
(4) exp03:data are limited over the Western Pacific region,
