A Brief Introduction of a Realtime Weather-Adaptive 3DVAR Analysis System

With Automatic Storm Positioning and On-demand Capability

 

Jidong Gao

A real-time, dynamically-adaptive 3DVAR system that incorporates all available radar observations was developed recently (Gao et al. 2009, Stensrud et al. 2010). A storm-positioning program is implemented in the 3DVAR system to automatically locate three 200 km x 200 km analysis domains in regions of thunderstorm activity based on the NSSL WDSS-II composite reflectivity product. The analysis can also be performed with on-demand capability in which end-users (or forecasters) specify the location of the analysis domain in real time based on the current weather situation in the fourth analysis domain. The 3DVAR system has the ability to automatically detect and analyze severe local hazardous weather by identifying mesocyclones at high spatial resolution (1km horizontal resolution) and high time frequency (every 5 minutes) using data primarily from the national WSR-88D radar network. The 3DVAR analyses contain the full three-dimensional wind field and hydrometeor fields, providing information beyond that of the radar observations alone. In addition, by combining observations from multiple radars, the 3DVAR system provides a single information source that can reduce the observational data flow that challenges forecasters every day. The performance of the system in the spring of 2010 is documented in Stensrud et al. (2010) and Smith et al. (2010). The system will be continuously tested in spring of 2011. The preliminary webpage which demonstrates the system is located at:

http://www.nssl.noaa.gov/users/jgao/public_html/analysis.

User Instructions:

1)     You can view any available product by choosing year, month and date. After choosing the date, the first line of left table is the general list. To make sure you can always view the current updated analysis, you need frequently click the button Refresh_Today_Product after a while. To return to the general list of the product, you may click List_All_Product.

2)     The second line Current Location includes four analysis domains indicated by 00, 01, 02, 03. The first three domains are weather-adaptive domains. The first one 00 is the most possible severe weather analysis domain, then 01 second possible, 02 the third possible. The 03 domain can be set up by the end users (forecasters).

3)     For the Current Types, 01 displays fields (horizontal vorticity, wind vectors and reflectivity) at 3 km Above Ground Level (AGL); 02 displays the fields at 1.5 km AGL; and 03 displays vertical slices of the fields through maximum vertical velocity.

4)     To view the movies (every half hour), you need click buttons below hour, to view individual fields at particular time level, you need click the buttons below minute.

5)     To switch domain, you need stop at any minute; then click domain button 00, 01, 02, or 03; after that, come back to any hour, or minute buttons.

6)     To switch different slices, you need click type buttons 01, 02, or 03, then come back to any hour, or minute buttons. The other numbers 04, 05, 06 are reserved for future use.

This is a very preliminary web display; if you have any questions, please let me know (email to: jidong.gao@noaa.gov).

 

References:

Gao, J., D. J. Stensrud, and M. Xue 2009: Three-dimensional Analyses of Several Thunderstorms observed during VORTEX2 field operations. 34th Conference on Radar Meteorology, Willimsburg, VA., Online publication.

Smith, T. M. Smith1, K.M. Kuhlman, K. L. Ortega1, K. L. Manross, D. W. Burgess, J.Gao, and D. J. Stensrud 2010: A survey of real-time 3DVAR analyses conducted during the 2010 experimental warning program spring experiment. Extended Abstract, 25th Conference on Severe Local Storms, Denver, Colorado.

Stensrud, D. J., J. Gao, T. M. Smith, K. Manross, J. Brogden, and V. Lakshmanan, 2010: A realtime weather-adaptive 3DVAR analysis system with automatic storm positioning and on-demand capability. Abstract, 25th Conference on Severe Local Storms, Denver, Colorado.