Status: Finalised Material from: Fernando, David L., Linus
Picture
1. Impact
A major storm was responsible for the floods in California on 17-19 February (strong winds were reported as well). This was not the first event of the season and California has been battered by anomalous wet weather. The week before more than 150K people were forced to abandon their homes due to a problem with a crack in Lake Oroville's dam as a result of excessive water. The anomalous wet season came after several years of drought conditions in California.
2. Description of the event
The extreme event was associated with a change of the weather pattern. The sequence of MSLP & 200hPa winds maps below are valid between 14 to 19 February shows the high pressure system, located NE Pacific, retreating (& weakening) westwards allowing the jet and disturbances to move towards the West Coast. A rapid cyclogenesis is taking place steered by the upper level jet as it moves SE towards the Baja California.
Daily totals of "observed" rainfall are available at the NWS (NOAA) web site. The word is in quotes because the maps were generated from the rain gauge "against long term climatologic precipitation (PRISM data), and derived amounts are interpolated between gauge locations" in mountain areas, which seems to be this case. The 3 maps show the rain pattern sliding down the coast of California.
The totals (24h) reported by the rain gauges for the 18@12 in south of California are displayed below (inch). They seem to be consistent with the totals generated for the gridded map.
Maps of total precipitable water based on an algorithm using microwave observations from polar orbiting satellites that are generated at Cooperative Institute for Meteorological Satellite Studies (CIMSS) are available to the public. Based on these maps we can detect the presence of atmospheric rivers (ARs) which were responsible for the large amounts of precipitation that occurred in California last weekend (see DR of yesterday). A stream of high values of precipitable water stretches from subtropical region, crossing the Hawaiian Islands towards the coast of Washington State, sliding equatorwards along the coast and reaching Southern California 2 days later, embedded in the extra-tropical cyclonic circulation.
The development of this AR in the central subtropical Pacific seems to coincide with the propagation of a strong MJO event in the western equatorial Pacific some days before. Looking at a few products that monitor the presence and progression of the MJO we can see that around 10 February the system, in terms of OLR anomalies was located in the western Pacific (top panel). This is also confirmed by the Wheeler and Hendon index based on the forecast initialised on the 9th February (left panel). The extended range forecast initialised on the 26th January shows the progression of an MJO event in Indian Ocean. By day 20 (~15 January) a good fraction of the ENS members predict a strong MJO in the western Pacific.
The rainfall during the winter has put an end to the drought in California. The plots below show that drought monitoring maps from February 2014, 2015, 2016 and 2017.
The plot below shows the accumulated precipitation for 8 stations in northern California from http://cdec.water.ca.gov/cdecapp/precipapp/get8SIPrecipIndex.action
3. Predictability
3.1 Data assimilation
3.2 HRES
3.3 ENS
The EFI signal of abnormal rainfall over the region 7 days before the event (and SOT values indicating an extreme scenario). As we get closer to the event the EFI become higher.
The atmospheric river event behind the extreme rainfall in some parts of California was indeed an abnormal case (see above). David Lavers has been working on an EFI of the "vertically integrated horizontal water vapour flux", as shown in the maps below (note the title on the maps is incorrect). The maps show the EFI on forecast day 10 (left plot) and forecast day 1 (right plot), both valid for the 15th February. An impressive large scale plume of v. integrated horizontal water vapour flux dominates the North Pacific region (in fact, the largest pattern seen at a global scale). As expected the signal is stronger as we get closer to the event, but it is remarkable how well the amplitude of the EFI, the extension and phase (location) are captured ten days before.
3.4 Monthly forecasts
The series of plot below show weekly precipitation anomalies for 13-19 February (note that the California event was at the end of the week).
As the winter 2016-2017 was very wet, it is relevant to look at the seasonal forecast. However, there is very little trace of rainfall above normal in south-western U.S in the forecast from 1 November for DJF.
