Prof. Avelino F. Arellano (January 20, 2022, 09:00-10:30)

Affiliation University of Arizona, USA
Title The Impact of Precipitable Water Vapor Data Assimilation on North American Monsoon Precipitation Forecast in Northern Mexico

While precipitation during the North American Monsoon (NAM) significantly contributes to the hydrometeorology of the southwest United States and northwest Mexico, current numerical weather prediction (NWP) systems lack the necessary skill in forecasting NAM precipitation, especially on events associated with weak synoptic forcing. This is in part due to the lack of relevant observational constraints in this semi-arid to a rid region that is mostly influenced by complex topography. Here, we present a couple of observational, modeling, and assimilation activities that we have been developing over the past 5 years or so to improve short-term NWP forecasts of monsoon precipitation in this region. In particular, we use a custom-built NWP system consisting of: a) precipitable water vapor retrievals taken from a Global Navigation Satellite System (GNSS) ground-based GPS network that my colleagues have deployed during summer of 2013 and 2017 over northwest Mexico; b) convective-permitting (CPM) Advanced Research Weather and Forecasting Model (WRF-ARW) with NCEP GFS (and NAM) as initial and lateral boundary conditions (IC/BCs); and c) an ensemble adjustment Kalman Filter (EAKF) implemented in the NCAR Data Assimilation Research Testbed (DART).
We assimilate only GPS precipitable water vapor (GPS-PWV) from 18 sites in forecast simulations of 38 weakly forced days during the 2017 NAM season. We use two 12-hour data assimilation (DA) cycle with one- hour interval assimilation, and 30 ensemble members. Single deterministic forecasts are initialized at 18 UTC and ending at 12 UTC the following day. We then compare these forecasts with those simulations initialized with no DA (NODA). Model precipitation is evaluated against Level-3 precipitation products from the Integrated Multi-satelliteE Retrievals for Global Precipitation Measurement Final product (GPM IMERG Final) while model cloud top temperature (CTT) is verified using infrared products from GPM MERGIR.
Our results show that the assimilation of the GPS-PWV improves the initial condition of WRF by reducing the root-mean-square error and bias of PWV across 1200-1800 UTC. This also leads to an improvement in capturing nocturnal convection of mesoscale convective systems (MCSs; after 0300 UTC) and to an increase by 0.1 mm per hour in subsequent precipitation during the 0300-0600 UTC period relative to no assimilation of the GPS-PWV (NODA) over the area with relatively more observation sites. Moreover, we find that the GPS-PWV DA decreases CTT, increases most unstable convective available energy (MUCAPE) and surface dewpoint temperature, and thus creates a more favorable condition for convective organization in the region. This work serves to establish a methodological approach in assimilating GPS-PWV (and data from opportunistic observing system of moisture from commercial microwave links) in short-term CPM forecasts in other areas of the world. Many countries in the subtropics and tropics face similar logistical challenges as Mexico, with respect to availability of high spatiotemporal resolution observational meteorological data to initialize high- resolution NWP forecast systems.

Back to Seminar Series

Page Top