An external view of the Met Office building at night.

Dr Zhihong Li

Zhihong works on data assimilation methods for 1.5 km resolution NWP-based nowcasting systems.

Areas of expertise

Zhihong's areas of expertise include:

  • Convective-scale NWP-based nowcasting system
  • 3-dimensional and 4-dimensional variational analysis (3D-Var and 4D-Var)
  • Nudging
  • Coastal and shelf seas modelling

Current activities

Zhihong is a senior scientist working on 3D-Var and 4D-Var methods for convective-scale resolution versions of the Unified Model in a NWP-based nowcasting system for forecasts on timescales of 0–6 hours ahead.

Nowcasting systems have traditionally used methods of localised analysis and extrapolation of observational fields such as cloud and precipitations. Such approaches have however a very limited capability of forecasting any developing weather system, which often causes hazardous weather conditions including severe flood events.

The accuracy of very short-range NWP depends critically upon the accuracy of the initial conditions and the ability of the model to retain information derived from frequent high resolution observations.

Zhihong's work is focusing on the development and testing of methods based on the Met Office's incremental 3D-Var and 4D-Var in high-resolution versions of the Unified Model, with a special emphasis on verification of rainfall, fog, and cloud.

A rapid hourly data assimilation cycle is being tested to assimilate high temporal and spatial novel observations such as radar Doppler winds, radar reflectivities into the 1.5 km model.

Career background

Zhihong joined the Met Office in 1999 in Ocean Forecasting Research (OFR), having previously spent 10 years as a research scientist specialising in coastal and shelf seas numerical modelling at University of Wales Bangor. In OFR, Zhihong's work included how to make best use of atmospheric forcing to the ocean model, and how ocean surface parameters (sea-surface temperature and currents) affect the NWP model predictions. Zhihong has been a member of the JCMM Data Assimilation group since 2003, involved in the development of 3D-Var assimilation system for the Met Office's initial high-resolution forecasting trial system, which led to implementation of operational UK 4 km and 1.5 km models.