Crawfish Tales

A Quarterly Publication of the National Weather Service Lower Mississippi River Forecast Center
Slidell, Louisiana

David Reed, Hydrologist In Charge Vol. 3No. 2, Winter 2000 Ethan A. Jolly, Editor

Features Radar Vs. Rain Gage Data XSETS

From the HIC

LMRFC started the new year with no Y2K problems with our major applications and our ability to prepare and disseminate forecasts was uninterrupted. With this hurdle in our commmunications and processing systems over, we are now ready to complete the final work to commission the AWIPS system at LMRFC and all the NWS. With this commissioning comes the decommissioning of our old AFOS system and its removal by the end of this fiscal year.


During the past quarter, the LMRFC welcomed two new employees, Amanda Roberts and David Welch, to the staff. Both have masters degrees in Geology and have a background in hydrology. There training is ongoing and they will be prepared to stand shift by the summer. Welcome to David and Amanda!


The four RFCs in Southern Region will be participating in a test of routine 24-hour operations. The test will start February 13 and end December 2. During the June 18 through July 15 period, 24-hour operations will be suspended and then resume on July 16. A 14 member evaluation team composed of NWS field and headquarters personnel and representatives of our users has been formed to conduct the test and evaluate the results. Our cooperators will be contacted by members of the evaluation team as part of the evaluation. The team will evaluate the importance of and need for routine 24-hour operations at an RFC and prepare a written report. Thank you for all comments provided to the team.

A happy and prosperous year 2000 to all our users and customers!

- Dave Reed

Radar Vs. Rain Gage Data

In a typical year flooding kills more people in the United States than does any other weather phenomenon. Major efforts have been underway for several years to improve flood forecasts. For example, technology has advanced the skill of river forecasting by allowing streamflow models to be run on local computers at river forecast centers. The shortage of observations from sparse rain gage networks has been a limitation; however, today's forecasters have more available data, including observations at both a greater spatial and temporal resolution. Forecasters now are beginning to test fine scale radar-derived rainfall data

which could prove to be a great advantage in river forecasting.


The LMRFC is examining the differences and similarities between mean areal precipitation derived from radar (MAPX) to that derived from rain gages (MAP). Initial studies were conducted for the Flint River in central GA, specifically the Flint River at Culloden (Stellman et al, 1999). The study has been expanded to 13 basins across the Lower Mississippi River Forecast Center area. The areas in this study are shown in figure 1.

Figure 1. Thirteen basins within the LMRFC area studied

A composite of radar-derived rainfall estimates and rain gage data on a 4 x 4 km grid at hourly intervals was used to generate a time series of basin MAPX at 6-hour intervals. MAP was computed using the Thiessen Polygon Method from daily and hourly rain gages.

For the areas in figure 1, MAPX was compared to MAP for 2 1/2 summer seasons (June-August) when convective activity is predominant, and 2 winter seasons (November-March) when stratiform precipitation is expected, as well as for the entire period (June 96-July 98). For the basin in Central GA, MAPX underestimated rainfall by about 50% during the winter months (Fig. 2). However, the correlation between MAP and MAPX is greatest during these months implying that the radars capture the precipitation but underestimate it. When comparing the two during the summer season, the MAPX means are nearly equal to MAP, but the correlations are lower (Fig. 2). These results are expected during summer due to the sparcity of the rain gage network and the spatial variability of convective events.

Figure 2: Mean MAPX and MAP by month for CLUG1 during periods that reported rainfall. Also shown is correlation, standard deviation, and number of 6 hour rain periods.

Values of MAPX and MAP were summed over the entire period (June 96-July 98 for CLUG1 and Sept 96-June 99 for the other 13) to assess any differences in total rainfall. MAPX underestimates rainfall by about 40 inches for the two-year period compared to the individual gages and MAP for CLUG1 and about 60 inches for MRML1 in southwest LA (Fig 3). Since MAPX and MAP values are similar during the summer, the deficit arises during the winter season. These kinds

of discrepancies would have large adverse effects on a model simulating streamflow.


All but one of the 13 basins in the LMRFC area showed that MAPX underestimates MAP with significant emphasis on the winter season. The lone basin which showed similar rainfall amounts between MAP and MAPX is located in northern AR where a dense network of rain gages may have helped reduce any overestimate of MAP that results from a sparse network as well as improve the Stage III estimate.

Figure 3: Cumulative sum of MAPX and MAP for Mermentau, LA (top) and Culloden(bottom), along with a daily rain gage site (Talbotton), an hourly rain gage site ( ATL-NWS) .

For radar derived precipitation, previous studies have shown that range effects are a common problem with underestimation (Fo, 1998). However, we believe that the Z-R relationship being used is the main source of error in the precipitation estimates largely due to variations in drop size distributions from storm to storm (Brandes et al., 1999). The Stage III process attempts to correct for these problems by using a bias computed from radar-gage pairs. However, there must be several gage-radar pairs for a new bias to be computed, at which time the bias is applied uniformly across the radar. This may cause problems because storms and the bias vary across the radar, especially during winter when convection and stratiform rainfall co-exist on a larger scale.

Future work includes comparing MAPX to a MAP derived from raw radar data to examine any improvements made by the Stage III process.


There are several recommendations for river forecasters based on the results of this study:

1.Monitor the radar bias to be sure that the same bias is not used for extended periods,

2.Increase the area of influence of rain gages during winter when stratiform rainfall is common,

3.Use MAPX for rainfall distribution in a qualitative manner, since we know that it is not quantitatively ready yet, and

4.Evaluate the calibration of soil moisture accounting models using radar data before the results are used in forecasting.


Brandes, E.A., J. Vivekanandan, and J.W. Wilson, 1999: A comparison of radar reflectivity estimates of rainfall from collocated radars. J. Atmos. Ocean. Tech., 16, accepted for publication

Fo, A.J.P., K.C. Crawford, and C.L. Hartzell, 1998: Improving WSR-88D hourly rainfall estimates. Wea. Forecasting, 13, 1016-1028.

Stellman, K., H. E. Fuelberg, and R. Garza, 1999: Utilizing radar data and semi-distributed modeling in streamflow simulations. Florida State University, pp. 78.

- Keith Stellman

XSETS - the new standard for river forecast products

The Lower Mississippi River Forecast Center will make AWIPS its official operating and communication system in the middle of April 2000. This achievement also marks the end of an era for AFOS, the communication system that has been in operation in the National Weather Service for the past 20 years. The LMRFC, along with many NWS and other government agency personnel have been diligently working together to bring all LMRFC operations and communications exclusively functioning on the AWIPS system. A test period will be initiated in the middle of February 2000 in which the LMRFC will begin using AWIPS exclusively for two months on a test basis. Not only at this point does all communications between LMRFC and other NWS offices have to be solely accomplished using AWIPS, so does all communications to and from the LMRFC and its external users. After testing is completed and is successful, AWIPS will be commissioned and AFOS decommissioned.

With AFOS, communication between NWS offices were limited to a 9600 baud rate. With AWIPS, communications between offices will be accomplished through a wide area network(WAN) with T1 capabilities. The differences between the two is like comparing a one lane dirt road to a 16 lane superhighway. Also most incoming products to National Weather Service offices will be over a satellite broadcast network in addition to the land linked WAN.


Official commissioning has already begun at a few NWS offices. AWIPS at all NWS offices should be commissioned by the end of August 2000.


- Eric Jones is the U.S. government's official web portal to all federal, state and local government web resources and services.