Passive Microwave Forward Modeling and Ensemble-Based Data Assimilation within a Regional-Scale Tropical Cyclone Model
Open Access
- Author:
- Sieron, Scott Buku
- Graduate Program:
- Meteorology
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- November 14, 2019
- Committee Members:
- Eugene Edmund Clothiaux, Dissertation Advisor/Co-Advisor
Eugene Edmund Clothiaux, Committee Chair/Co-Chair
Matthew Robert Kumjian, Committee Member
Xingchao Chen, Committee Member
Jia Li, Outside Member
David Jonathan Stensrud, Program Head/Chair - Keywords:
- Community Radiative Transfer Model (CRTM)
Weather Research and Forecasting (WRF)
hurricanes
tropical cyclones
data assimilation
microwave
infrared
satellite - Abstract:
- Passive microwave (PMW) observations are informative of liquid and ice water contents, qualities which well-characterize tropical cyclones (TCs). Their successful assimilation with weather forecast models could help improve TC forecasts. The Community Radiative Transfer Model (CRTM) forward model’s representation of clouds mischaracterize certain differing assumptions between microphysics schemes (MPSs) and may also be overall inconsistent with many MPSs. Three forecasts using the Weather Research and Forecasting model, each using a different MPS, are used with the CRTM to simulate PMW brightness temperatures (BTs). Cloud scattering look-up tables (LUTs) are constructed for consistency with each scheme. The custom and default LUTs lead to vastly different BTs from each other and to observations. There are cold biases across the MW spectrum, and BTs at 183 GHz are warmer than at ~90 GHz. New lookup tables for the WSM6 scheme are produced in which non-spherical particles replace the soft spheres implied by MPSs. Sector snowflakes for snow best resolves the issue of relatively warm 183 GHz BTs, but worsens the overall cold bias at high frequencies. Cold biases at lower frequencies caused by graupel could not be physically resolved. However, PMW BTs simulated from the mean state of ensemble members match well to observations, suggesting that ensemble Kalman filter (EnKF) assimilation could succeed. This hypothesis is tested using a regional-scale TC model. Infrared (IR) and conventional observations are assimilated, along with ~19-GHz and 183.31±6.6-GHz PMW observations, representing liquid and ice water contents, respectively. IR is assimilated when PMW observations are not available, otherwise different combinations of PMW and IR assimilation were tested. The precipitation structures implied by PMW BTs in both the EnKF analysis and subsequent short-term forecasts match better to observations versus assimilating only IR. PMW assimilation leads to significant and intuitive increments to thermodynamic and dynamic variables, and does not significantly degrade intensity and track forecasts from assimilating only IR. PMW assimilation may be more successful with further changes to data assimilation procedures, the application of the CRTM, and MW cloud scattering properties. Additional experiments may inform on the best procedures and be more conclusive of the impacts of PMW assimilation.