We consider the Incremental Strong constraint 4D VARiational (IS4dvar) algorithm for data assimilation implemented in ROMS with the aim to study its performance in terms of strong scaling scalability on computing arch...
详细信息
ISBN:
(纸本)9783319780542;9783319780535
We consider the Incremental Strong constraint 4D VARiational (IS4dvar) algorithm for data assimilation implemented in ROMS with the aim to study its performance in terms of strong scaling scalability on computing architectures such as a cluster of CPUs. We consider realistic test cases with data collected in enclosed and semi enclosed seas, namely, Caspian sea, West Africa/Angola, as well as data collected into the California bay. The computing architecture we use is currently available at Imperial College London. The analysis allows us to highlight that the ROMS-IS4dvar performance on emerging architectures depends on a deep relation among the problems size, the domain decomposition approach and the computing architecture characteristics.
Study region: The Eastern region of the Qinghai-Tibet Plateau (EQTP) Study focus: A regional numerical weather prediction and data assimilation system is constructed to investigate the impact of assimilating Global Pr...
详细信息
Study region: The Eastern region of the Qinghai-Tibet Plateau (EQTP) Study focus: A regional numerical weather prediction and data assimilation system is constructed to investigate the impact of assimilating Global Precipitation Measurement (GPM) precipitation and Himawari-8/Advanced Himawari Imager (AHI) water vapor radiance using Weather Research and forecast (WRF) model and Four-dimensional variational assimilation (4dvar) method on snow properties predictions. The predictions were compared with some reference datasets, including MODIS,VIIRS,GLDAS and ERA5-land. New hydrological insights for the region: DA_G&A showed a significant increase in deep snow area (SD >15 cm), and a decrease in shallow snow area (SD<5 cm). Comparing with some reference datasets, the predictions exhibit good physical consistency between snow parameters and fine temporal-spatial resolution. The forecasts are found to be reliable and reasonable. However, Noah-MP coupled in WRF tends to overestimate SCF and SAL, which is largely attributed to the limitations of the associated parameterization schemes. These findings highlight the assimilation of atmospheric data can improve the forecasting of snow properties. However, in Noah-MP, there remains significant uncertainty in the snow-related parameterization schemes and initial conditions.
A new variational formulation and a method for solving the problem of quasi-geostrophic dynamics in a two-layer periodic channel are considered. The area imitates the zone of the Antarctic Circumpolar Current lying in...
详细信息
A new variational formulation and a method for solving the problem of quasi-geostrophic dynamics in a two-layer periodic channel are considered. The area imitates the zone of the Antarctic Circumpolar Current lying in the Southern Ocean. One feature of the problem is the doubly connected region of its solution (periodicity in latitude). Using the expansion in Fourier series, the problem is reduced to a nonlinear system of ordinary differential equations (ODE) in time. The doubly connected region leads to the fact that, together with the ODE solution, it is required to satisfy the stationary integral relation that determines the total flow transport. A variational numerical algorithm for solving the problem is proposed which is close to the technique of four-dimensional data assimilation (4dvar). The basis of the cost function is the stationary integral relation. With the help of a series of computational experiments, the stationary regimes of flows depending on model parameters are studied. Calculations show that the presence of high harmonics in the bottom relief can cause the formation of a undercurrent in the lower layer. The undercurrent is stable to small variations in relief disturbances and the turbulent viscosity coefficient.
In this study, the investigation is made to reveal the impact of multi-strategically assimilating Global Precipitation Measurement (GPM) precipitation and Himawari-8/Advanced Himawari Imager (AHI) water vapor radiance...
详细信息
In this study, the investigation is made to reveal the impact of multi-strategically assimilating Global Precipitation Measurement (GPM) precipitation and Himawari-8/Advanced Himawari Imager (AHI) water vapor radiances (WVR) on forecasting a heavy snowfall event in the Eastern Qinghai-Tibet Plateau (EQTP) employing the Weather Research and Forecast model (WRF) and the Four-Dimensional Variational (4dvar) assimilation system (WRF-4dvar). The multiple data assimilation (DA) strategies include control tests (CON), the individual assimilation of AHI and GPM tests (DA_AHI and DA_GPM) and the joint assimilation of GPM and AHI (DA_G&A), &A), with different initial times. The results indicate that GPM precipitation effectively captures mesoscale atmospheric details, but its scope is confined to a limited area. AHI WVR is sensitive to upper-middle atmospheric humidity and furnishes extensive-scale environmental parameters such as water vapor transport characteristics. The joint assimilation of the two not only yields multi-dimensional atmospheric insights but also addresses the limitations of individual assimilation. Assimilation GPM and AHI are respective sensitivity to the lower layers (about 800hpa) and upper layers (about 400hpa) of model. The individual assimilation GPM has the greatest effect on near-surface humidity field, and AHI plays a dominant role in the joint assimilation. By assimilating different remote sensing products at different initial times of NWPs, the thermodynamic and dynamic structures are variously reconstructed, leading to the different snowfall scenes. In addition, we further compare the 12- hourly cumulative snowfall with in-situ meteorological station observations. The predictions of snowfall from DA_G&A &A perform much better with the correlation coefficient (CC) and root-mean-square error (RMSE) 0.36 and 3.14 mm, respectively. As for different initial times of NWPs, the best snowfall forecast is 0600 UTC on October 28, 2022, and the CC is 0.4.
暂无评论