#TITLE_ALTERNATIVE#

• Main Author: Joko Trilaksono (NIM 224 05 015), Nurjanna
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/6568
• Institution: Institut Teknologi Bandung

Abstract: <br /> <br /> <br /> <br /> <br /> The diurnal patterns of convection over Sumatra Island have been investigated from geostationary satellite data (GMS 5, GOES 9 and MT-SAT) at infrared channel (IR1) and numerical weather prediction model simulations. The Weather Research and Forecasting-Advanced Research WRF (WRF-ARW) developed and maintained by the Mesoscale and Microscale Meteorology Division of NCAR, which consist of Euler non-hydrostatic and fully compressible set of equation, has been used in the numerical simulation. Both initial and boundary condition for the simulations are taken from Global Tropospheric Analysis data operated by NCEP. <br /> <br /> <br /> <br /> <br /> <br /> From the satellite data, four periods of different convection scales over Sumatra Island has been identified based on Outgoing Longwave Radiation (OLR) data. Cloud top temperature composites (derived from IR1 images) from 2002-2006 suggest that there are temporal and spatial shifts for each convection scale, particularly for the phase of convection peak over land. The spatial shifts seem to indicate the important role of small islands along the western part of Sumatra Island. Moreover, the detailed daily convective activities over Sumatra Island show stronger random pattern. <br /> <br /> <br /> <br /> <br /> <br /> The numerical model simulations have been used to investigate diurnal patterns of convection for several days in November 2006 during the period of large-scale convection. The model output consistency test has been carried out to choose the best combination of parameterizations in this study. From several experiments, it is found that a combination involving YSU microphysics and MRF planetary boundary layer parameterization give the best result. Using this combination, the model successfully simulated diurnal convection peak over the land, which was indicated by correlation coefficient of cloud top temperature over 0.6. However, the peak phase over the sea has not been yet well simulated. Further investigation show that the local data (radiosonde) availability affects the model performance. These results indicate that the model is capable to take effects of complex topography of Sumatra Island into account but further study is needed to overcome the weakness of the model in simulating the convective phase in the sea. <br />