Research & Development Division

Specialized Medium Range Weather Forecasting Centre

Project brief

Pakistan Meteorological Department has initiated the step to enhance weather forecasting and warning capabilities with the help of JICA/Government of Japan. Under the project “Establishment of Specialized Medium Range weather Forecasting Centre (SMRFC) and Strengthening of Weather Forecasting System in the Islamic Republic of Pakistan” sponsored by JICA, installation of weather radar and a Wind Profiler system, High Performance Computing Cluster system has been installed at PMD Islamabad. The main objective of this project is to make available medium range weather forecasts (country wide, 3-10 days) using higher performance computing system. It will be capable for timely dissemination of accurate severe weather forecast and warnings to the vulnerable districts/communities of Pakistan.It will also be able to measure highly accurate rainfall intensity and to detect the weather situation.

Latest updates

Under Construction

Marine Weather Forecast

Model Output

Short Range Forecast

SMRFC project provides 0-3 days weather forecast. The products of following models are available:

  • WRF (Spatial resolution 5Km)
  • ICON (Spatial resolution 13Km)
  • JMA (Spatial resolution 20Km)
Models Output

Medium Range Forecast

SMRFC project provides upto 10 days weather forecast. The products of following models are available:

  • W557 (Spatial resolution 20Km)
  • GFS (Spatial resolution 25Km)
  • CMC (Spatial resolution 110Km)
Models Output

Flood Forecast

High resolution (5Km) WRF model products are avaialable for the flood catchment areas of Pakistan. The basins covered are Indus, Jhelum, Kabul, Chenab, Sutlej and Ravi River Basin

Model Output
OCF Operational Consensus Forecasts
Models Output

Operational Consensus Forecasts (OCF) employs multiple linear regression to develop forecasts using spatiotemporal interpolation of model values of surface and upper air variables (as potential predictors) and site observations (as predictands) in the dependent data set.

Basic features of OCF are as follows:

  • Use forecasts interpolated from > 5 numerical models.
  • Bias-correct each of the forecast series over a certain previous number of days.
  • Calculates errors of each bias-corrected forecast series.
  • Select the smallest set of n ‘medians’ and use their inverse as weights.
  • To make new OCF, normalise weights and multiply by corresponding bias-corrected forecasts.
  • Point Based – Observatory level forecast only
Note

The model products will be updated on daily basis.

The model products will be upgraded/replaced based on observation, on regular basis.

For any suggestion please contact the focal person.