The National HAS Conference was held in Silver Spring, Maryland, September 14th - September 19th, 1997. HAS (Hydrometeorological Analysis and Support) Forecasters from each of the thirteen river forecast centers attended the conference to discuss scientific and technical issues as well as some of the software tools that they use in their effort to assimilate meteorological data into the River Forecast models.

The participants from left to right

Seated: Greg Hanson (OSF), Jeff Smith (CBRFC), Tyree Wilde (NWRFC), Dave Cokely (MBRFC), Bill Lawrence (ABRFC).

Standing: Patty Wnek (MARFC), Owen Rhea (CNRFC), Jeff Graschel (LMRFC), Jeff Perry (AKRFC), Bob Wavrin (NCRFC), Judy Bradberry (SERFC), Althea Austin-Smith (NERFC), Mark Fenbers (OHRFC), Greg Story (WGRFC), Matt Lorentson (SERFC), Jay Breidenbach (OH/HRL).

Summary of Conference

The conference began with the Director of OH, Danny Fread, giving a management level overview of the end-to-end forecasting process, stressing how important the HAS function is to the entire concept. Dr. Fread emphasized that the HAS forecasters will play an increasingly important role in the assimilation of QPF and Probabillistic QPF to ultimately produce Probabillistic River Stage Forecasts. To adequately perform this function, a large amount of computer software, much of it yet to be written, will be needed.

Each RFC was given an opportunity to discuss the HAS function and software that is used at the RFC. From these discussions it immediately became clear that there has already been an enormous amount of local software developed for the specific needs of each RFC. It also was clear from these briefings that each RFC has a unique set of problems that they must be able to handle which varies significantly across the United States. The wide variation in climate, terrain, and precipitation characteristics among the RFCs has led to differences in the way the HAS function is run and software that has been developed. Despite these large differences there has been some duplication of effort. This conference has helped to clarify what data and tools are available locally, nationally, and what is being planned.

It was pointed out that some RFC's use Stage III or P1 radar/gage estimates directly in the model and others do not depending the circumstances of that particular RFC. There were also differences in software used to specify and mosaic QPF. RFC's east of the Rocky mountains primarily used HAS/WIN QPF or some slight variation. Most Western RFC's were using Mountain Mapper, developed at CBRFC for their QPF needs as well as Data QC, Precipitation Estimation, and Verification. The following table is an attempt to list some of the differences in software applications and data used by the HAS forecasters at the various RFC's. This list only reflects items that were discussed at the HAS conference.

RFC	RADAR data used in river model?	QPF Software	Other Useful Software /Data
OHRFC	Stage III (auto + qc in morning) used in model for some test basins	HAS/WIN QPF	GMT Gostscript Image Magic Various Scripts
MARFC	Stage III (auto) (qualitatively)	HAS/WIN QPF	ATLAS GIS - Mapping Software to display Gage Data, QPF, FFG etc
NERFC	Stage III (auto + qc where needed) (qualitatively)	HAS/WIN QPF	Locally Developed TCL/TK script used as interface to HAS QPF.
SERFC	Stage III (qualitatively)	HAS/WIN QPF	Atlas GIS.
WGRFC	Stage III (auto + manual QC) used in most basins	HAS/WIN QPF	Brainmaker - Neural Network QPF software.
LMRFC	Stage III (auto + manual QC) used in some basins	HAS/WIN QPF	Basin Average Rainfall Climatologies.
NCRFC	Stage III (Auto + maunual QC in morning) used in some basins (Warm Season Only)	HAS/WIN QPF	Home grown Software for Temperature Specification and Modification. Snow Water Content from NOHRSC.
MBRFC	Stage III used in forecast models (Warm Season Only)	HAS/WIN QPF	Various Scripts to process ASCII QPF files directly from WFOs.
ABRFC	P1 or Stage III used in all Basins	HAS/WIN QPF	XNAV - Displays just about everything.
NCRFC	Stage III and Mountain Mapper HDP mosaics used qualitatively and for gage QC	Mountain Mapper	Graphical HMD (see web site)
CNRFC		Mountain Mapper	Rhea Orographic Model
NWRFC	Stage III (qualitatively)	Mountain Mapper	Various Scripts
AKRFC			SAR images of River ICE

Since there are many useful scripts and software including many others not listed above that can be shared among the RFCs it was recommended that a location be established where software and scripts can shared by loading and unloading to and from a given location. The location was not defined but a logical place would be one that is accessible through the HRL home page.

Action Item: Establish location to share useful scripts and software

After the presentations from the RFC's, speakers from the Office of Hydrology, Office of Meteorology and the Operational Support Facility spoke on various topics. The following lists the topics and speakers and briefly summarizes important points and action items originating from the discussion.

Gage Networks -Ben Weiger

Gage Data Support System -Ed Johnson

• GDSS will be implemented at WGRFC, ABRFC and OHRFC for testing.
• Minimal impact on RFC operations.
• Ultimate location of GDSS will be AWIPS.

Procedures for obtaining more gage data -Mark Fenbers

• Most gages in OHRFC are state/county funded IFLOWS gages: however, there are many other source that could be used if work is undertaken to have the report SHEF encoded
• May require obtaining lat/lon through mapping software or GPS.
• Scan SHEF decoder log files for complaints then decide if they should be entered into eligibility table
• OHRFC uses team of staff members which regularly meets to deal with data issues

Proposed Changes to SIPP -Paul Tilles / Mark Fenbers

• Action item: Add an option in SIPP to create 6 and 24 hour reports from the 1 hour incremental gage reports

Data QC Issues -J.J. Pan / Geoff Bonnin / Heather Friedeman

• New data analysis and QC techniques have been developed for use with hydrologic data. They include the use of an automated double mass analysis tool, A multi scale time series analysis tool, and a cluster analysis tool.

Overview of Current QPF Specification Capabilities and Plans for Probabilistic QPF

-Tom Graziano

• Summary of AWIPS QPF meeting in Salt Lake City provided.
• Overview of plans for Probabilistic QPF.

Mountain Mapper -Jeff Smith

• Capability of Mountain Mapper demonstrated including Point QPF specification, rendering to gridded analysis using PRISM data, Daily QC using surrounding stations.
• Action item: OH to check with Tom Carrol on nationwide availability of PRISM data.

Ohio Floods Case Study -Glen Austin / Bryce Finnerty

• Impact of QPF shown to have some skill, however, magnitude and basin to basin variability was missed.
• Large QPFs must be used appropriately in Stream Forecast. (Does Hydrologist believe the forecast?)

Coordination Issues and Station Visits -Patti Wnek

• Stations visits to each of the WFOs by the HAS forecaster has fostered a greater cooperation between RFCs and WFOs (Puts a face with the voice on the phone)

• Important to keep a log of all of the coordination calls

Automated Satellite Derived Precipitation -Gilberto Vicente

• Summary of NESDIS effort to develop satellite precipitation estimates
• Action item: Incorporate Satellite precipitation estimates into Stage 2/3.

Operational Satellite Precipitation Estimates -Rich Borneman

• IFFA (Interactive Flash Flood Analyser) described.
• Will probably be replaced by the Vicente Automated Satellite Precipitation estimates

Training and Evaluation Modules: Changing Precipitation Adaptable Parameters for Tropical Situations

-Greg Hanson

• TEM's described as a useful training technique and a tool for SOO's and HAS forecasters to help train the WFO or RFC Staff on a particular issue such as use of appropriate Z-R relationships, etc. If a need for a TEM can be identified in your region or office, the OSF will work with you to write one.

Improving NAPUP Operations -Mark Albertelly

• Report all NAPUP operation problems directly to OSF. They will make sure the RPG is rebooted at a given site. This should free up the RFC from having to constantly ask trouble sites to perform the reboot.
• OSF is planning on setting up a multi-port communications monitoring system if resources permit.
• Action Item: HAS forecaster need to convince HIC's that monitoring by the OSF is important and have them push it through their respective regions.

Improving PPS estimates close to the Radar -Tim O'Bannon

• The new terrain following hybrid scan was described. It should significantly improve estimates close to the radar, especially at mountain top sites.
• Action Item: OH should push for the generation of the terrain following Hybrid Scan at all sites. (However, further evaluation is needed before national implementation)

Build 10.0 Enhancements to PPS -Dennis Miller

Planned PPS Enhancements in Open System -D.J. Seo / Dennis Miller

• Plans for the PPS in the Open Systems described. Initial focus to be on 1) new bias adjustment algorithm. 2) range correction algorithm. 3) bright band detection and mitigation.
• Action Item: RFC's need ability to request old products such as missing HDP's
• Action Item: RFC's would like ability to function as an associated PUP.

Steps towards regional reflectivity mosaics -David Kitzmiller

P1 -Bill Lawrence

• P1 described as providing superior estimates to Stage III in stratiform rainfall situations in heavily gages areas. Other features allow quicker interaction with the data for QC purposes, especially in stratiform or snow situations.

Differences between P1 and Stage III -Bill Lawrence / D.J. Seo

Recent and Planned Improvements to Stage II/III -D.J. Seo

• Action Item: Implement new bias adjustment algorithm which includes dual bias estimates (short term and climatological)
• Action Item: Implement RFC wide multisensor estimation as opposed to estimating rainfall for each radar and then mosaicking.
• Action Item: Implement new GUI features to allow quick interaction with the data much like the P1 algorithms. 1) Make Snow 2)Draw Polygon around and area and multiply by specified number (useful for bright band). 3) Ability to set all gages in a given area to missing. (Zero values are still a big problem in Stage III and P1. 4) Global Remove AP option as opposed to going radar by radar. 5) Ability to create a list of bad gages though a GUI. These gages would be set to missing until the user declared them to be good again.

Stage II/III Adaptable Parameters GUI -Jay Breidenbach / Mark Glaudemans

• Action Item: Implement ability for RFC to specify parameters for all the radars in the RFC area of responsibility.

Stage III Adaptable Parameter Optimization -Jay Breidenbach

• Action Item: HRL will provide updated guidance on adaptable parameters