DECEMBER
2007 FLOOD EVENT
PART 2: FORECAST VERIFICATION
December 2007 Event Verification
The verification discussion is divided into
two sections.In the first section the
primary meteorological input, quantitative precipitation forecast (QPF), is
evaluated.Evaluation of the hydrologic
forecasts follows.
Meteorological Verification
This was the first major event where hydrometeorologists at the NWRFC utilized GFE to generate
forecasted mean areal precipitation (FMAP) for input
to hydrologic models.Hydrometeorolgical forecasters started using GFE generated
QPF operationally this season.Each
morning the HAS forecaster initializes NWRFC's QPF
field using a mix of WFO ISC, HPC, and various model (GFS and
Presentation of QPF verification will start
off with spatial products generated during the event.This will be followed by basin specific
analysis of point forecasts.Figure 1 displays
the 24 hour QPF field in GFE for day 1.
The derived calibration point precipitation values for that same 24 hour
period are shown in figure 2.These can
be compared with the image of observed precipitation displayed on the
verification web page (see figure 3).

Figure 1 NWRFC GFE display ofday 1 QPF from

Figure 2 NWRFC point QPF plot for day 1 plot from

Figure 3 Observed cumulative precipitation from
In fig. 1 there is a 7.45 inch 24 hour
day-1 forecast accumulation sample value in the north
The day 2 GFE QPF grid shown in figure 4
showed that less precipitation was forecasted for Monday. The point QPF product
displayed in figure 5 shows that the day 1 GFE QPF grid generated Monday
morning (not shown) was an improvement over the forecast two days out.

Figure 4 NWRFC GFE display of day 2 QPF from

Figure 5NWRFC
point QPF plot for day 1 plot from

Figure 6Observed
cumulative precipitation from


Figures 7 and 8 Comparison of Precipitation Forecasts for
the


Figures 9 and 10Scatter Diagrams
of Observed Precipitationand NWRFC QPF
for the
|
|
||||||||||
|
Event Performance Statistics |
||||||||||
|
Event Performance -
MAJOR EVENTS |
|
|
Event Performance -
MAJOR EVENTS |
|||||||
|
1.00"+ in
6-hour Time Period |
|
|
4.00"+ in
24-hour Time Period |
|||||||
|
|
NWRFC |
HPC |
PQR |
|
|
|
NWRFC |
HPC |
PQR |
|
|
POD |
0.47 |
0.77 |
0.30 |
|
|
POD |
0.50 |
0.83 |
0.00 |
|
|
FAR |
0.18 |
0.08 |
0.10 |
|
|
FAR |
0.40 |
0.00 |
1.00 |
|
|
CSI |
0.42 |
0.72 |
0.29 |
|
|
CSI |
0.38 |
0.83 |
0.00 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Event Performance -
MODERATE EVENTS |
|
|
Event Performance -
MODERATE EVENTS |
|||||||
|
0.50"+ in 6-hour Time Period |
|
|
2.00"+ in 24-hour Time Period |
|||||||
|
|
NWRFC |
HPC |
PQR |
|
|
|
NWRFC |
HPC |
PQR |
|
|
POD |
0.90 |
0.93 |
0.71 |
|
|
POD |
1.00 |
0.92 |
0.75 |
|
|
FAR |
0.12 |
0.11 |
0.03 |
|
|
FAR |
0.00 |
0.00 |
0.00 |
|
|
CSI |
0.81 |
0.83 |
0.70 |
|
|
CSI |
1.00 |
0.92 |
0.75 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Event Performance -
MINOR EVENTS |
|
|
Event Performance -
MINOR EVENTS |
|||||||
|
0.25"+ in 6-hour Time Period |
|
|
1.00"+ in 24-hour Time Period |
|||||||
|
|
NWRFC |
HPC |
PQR |
|
|
|
NWRFC |
HPC |
PQR |
|
|
POD |
0.89 |
0.93 |
0.88 |
|
|
POD |
0.89 |
0.89 |
0.78 |
|
|
FAR |
0.09 |
0.07 |
0.09 |
|
|
FAR |
0.00 |
0.00 |
0.00 |
|
|
CSI |
0.82 |
0.87 |
0.81 |
|
|
CSI |
0.89 |
0.89 |
0.78 |
|
Table 1 Event QPF Verification
Statistics for the
|
|
||||||||||
|
Daily Forecast Performance Statistics |
||||||||||
|
Forecast Bias (Mean
Algebraic Error) |
|
|
Forecast Bias (Mean
Algebraic Error) |
|||||||
|
6-hour Time Period, All Stations |
|
|
24-hour Time Period, All Stations |
|||||||
|
|
NWRFC |
|||||||||