When an AR reaches land, its embed-ded
low-level jet stream strikes coastal
mountains, which forces it to ascend. As
it ascends, it cools, and large parts of its
vapor condenses. The resulting precip-itation
can be intense and wide-ranging
with long lasting effects. These impacts
can vary spatially depending upon the
orientation of the AR relative to the land
and other airflows. Changing winds in
ARs at landfall can cause them to stall
over certain areas and amplify their im-pacts.
(Neiman et al. 2016, Ralph et al.
2016, among others).
Forecasting
Recent research suggests that better
coupling of climate forecasts with sea-sonal
weather forecasts of ARs could help
improve water management as well as
decisions regarding flood control and wa-ter
supply operations. National Weath-er
Service forecasters located along the
West Coast of the United States are now
familiar with the concept of atmospheric
rivers and can identify these phenomena
in currently available numerical forecast
models. This provides the capability to
provide advanced warning of potential
heavy rain sometimes five to seven days
in advance. Forecasters have also learned
to monitor polar orbiting microwave
satellite imagery that provides informa-tion
on the presence and movement of
ARs in the Pacific. (ARs differ from other
weather patterns by the amount of inte-grated
water vapor in the atmosphere
over a large and narrow geographic ex-tent.
Cyclones and fronts are defined by
such characteristics as surface pressure
contours and changes in temperature
and winds.)
Filaments of enhanced water vapor
extending from the subtropics to the ex-tratropics
commonly observed in satellite
imagery are routinely used as a proxy for
identifying ARs. However, while the oc-currence
of an AR can be forecast up to
10 days out, and models provide useful
heads-up for AR impact and integrat-ed
water vapor (IWV) content, landfall
location, timing, and duration are less
well predicted. Location predictability
is subject to significant errors, especially
at longer lead times. The current ability
to predict the location of AR landfall in
three- to five-day forecasts is compara-ble
with current hurricane track error. Er
28 SOURCE winter 2017
November 2008 annotated NOAA GOES-11 infrared satellite image showing an atmospheric river making
landfall with a powerful cyclone. Effects in the northwest United States included significant rainfall,
snowfall, and waves more than 500 miles south of the storm’s low pressure center. Courtesy F. Martin Ralph.
rors in location increase to over 800 km
(roughly 500 miles) at 10 days out.
In 2013, California DWR began estab-lishing
a network of atmospheric river
observatories (AROs) (White et al. 2013).
These observatories make it possible for
forecasters to monitor the strength and lo-cation
of ARs as they make landfall along
the California coast and improve short-term
rainfall forecasts for flash flooding.
Each of these AROs includes a Doppler
wind profiler, a GPS-MET (GPS-Meteo-rology)
station that reports the IWV, and
surface meteorology. Thus equipped,
AROs can assess the incoming “fuel” for
precipitation events by measuring the
IWV flux coming on land. The IWV pro-vides
a measurement of how much water
is available in the atmospheric column,
and the flux represents how fast the col-umn
is moving (depth times speed).
Challenges remain in predicting rain-fall
totals in AR events because current
models aren’t equipped to project de-tails
of the duration and timing of ARs
as they make landfall. AR impacts are
enormous and spread over a large area.
For example, during one storm in Janu-ary
2008, major impacts were felt more
than 500 miles south of the low pressure
center of the storm. This was due to the
widespread nature of AR conditions as-sociated
with the cyclone and the amount
of moisture transported onshore. In addi-tion,
the precise characteristics of these
rivers are not operationally monitored
offshore. Field campaigns using air-craft
have flown into offshore ARs and
dropped weather balloons that sample
meteorological conditions, including the
vertical structure of water vapor content,
to better understand their precise evolu-tion
and “track.” These data have been
fed directly into weather models to help
improve AR forecasts.
Application of Forecast Data
The Center for Western Weather and
Water Extremes (CW3E) is working
with the Sonoma County Water Agency
(SCWA), the U.S. Army Corps of Engi-neers
(Army Corps), National Oceanic
and Atmospheric Administration (Na-tional
Weather Service, Office of Atmo-spheric
Research and National Marine
Fisheries Service), U.S. Geological Sur-vey,
U.S. Bureau of Reclamation, and
California DWR to research how mod-eling,
currently available forecasting
tools, and improved information might
be applied to improving AR-related
flood control and water supply manage-ment.
SCWA provides drinking water to
600,000 people in Sonoma and northern
Marin counties and is also responsible for
maintaining minimum in-stream flows in
the Russian River to meet water supply
needs of several communities and a large
agricultural industry along the Russian
River. In addition, SCWA must provide