...
The figures below show analyses of z500 and t850 (left) and 2-metre temperature and 30-metre wind (right) valid for 20 June 12 UTC.
The figure below shows the EFI for the mean temperature on 20 June. The high EFI values are fairly localized to the region around Karachi.
The plot below shows the time-series of observations (blue) for Karachi (25N, 67.2E) for 4 days. The HRES forecast from 19 June 00UTC is included (blue). The heatwave peaked on 20 June with 43 degrees and the minimum temperature the night after was 34 degrees.
3. Predictability
...
The HRES forecast is included in the observation plot above.
3.3 ENS
The plot below shows the Ensgram for Karachi from 20 June 12 UTC (the earlier Ensgrams had been removed when the investigation started). For the Sunday, both the daily maximum almost reaches the 99th percentile of the climate while the minimum temperature is above. We also see that the wind speed is much lower than normal.
The plots below show CDF for mean (left), max (middle) and minimum temperature for Karachi on 20 June. Both the minimum and maximum temperature seems too low compared to the observations, also for the last forecast (red line). However, local conditions around the measurement station could play a big role here.
3.4 Monthly forecasts
The plot below shows the verification of weekly temperature anomalies for the week 15 June to 21 June.
3.5 Comparison with other centres
...
5. Good and bad aspects of the forecasts for the event
- The heatwave seems to have been caused by a combination of high temperatures and weak winds. A heat index is probably needed to see the severity in the forecast.