River flow forecasting
and an upstream scheme in space are used for the
integration.
rainfall accumulations using 10-level model data. Meteorol.
Mag., 107: 113–124.
Bell, V. A. & Moore, R. J. (1998a). A grid-based distributed
flood forecasting model for use with weather radar data. 1.
Formulation. Hydrology and Earth System Sciences, 2:
Appendix C. Definition of performance
measures
2
65–281.
Bell, V. A. & Moore, R. J. (1998b). A grid-based distributed
flood forecasting model for use with weather radar data. 2.
Case studies. Hydrology and Earth System Sciences, 2:
The definition of rmse is:
0.5
2
83–298.
1
N
2
rmse =
(
Q − q )
Collier, C. G. (1975). A representation of the effects of
topography on surface rainfall within moving baroclinic
disturbances. Q. J. R. Meteorol. Soc., 101: 407–422.
n
n
∑
where Q and q are the observed and forecast value at Cullen, M. J. P. (1993). The unified forecast/climate model.
n
n
Meteorol. Mag., 122: 81–94.
the time n or the point n. The summation is computed
over N values in space or time.
Kessler, E. (1969). On the distribution and continuity of
water substance in atmospheric circulations. Meteorol.
Monographs, No.10, Am. Meteorol. Soc., 84 pp.
Lilly, D. K. (1990). Numerical prediction of thunderstorms –
has its time come? Q. J. R. Meteorol. Soc., 116: 779–798.
Miller, N. L. & Kim, J. (1996). Numerical prediction of pre-
cipitation and river flow over the Russian river watershed
during the January 1995 California storms. Bull. Am.
Meteorol. Soc., 77: 101–105.
The correlation coefficient, r , is defined as:
c
(
Q −Q )(q − q)
∑
n n
=
c
r
0
.5
0.5
2
2
[
Σ(Q −Q ) ] [Σ(q − q) ]
n n
—
–
where Q and q are the mean of the observations and Moore, R. J. & Bell, V. A. (1996). A grid-based flood fore-
forecasts, respectively, over the N points. Criterion r is
only used for evaluation of the rainfall distribution.
casting model using weather radar, digital terrain and
Landsat data. Quaderni Di Idronomia Montana, 16:
c
9
7–105.
Moore, R. J., Bell, V. A., Roberts, G. A. & Morris, D. G.
1994). Development of distributed flood forecasting mod-
els using weather radar and digital terrain data. R&D Note
52, Research Contractor: Institute of Hydrology, National
2
The goodness of fit, R , is defined as:
(
2
Σ(Qn − qn
)
2
R =1−
2
2
Σ(Q −Q )
n
Rivers Authority, 144 pp.
Moore, R. J., Carrington, D. S., Jones, D. A., Stewart, E. J.,
Hatton, R. & Aucott, L. (1994). The UK HYREX project.
Ann. Geophysicae, 12: Suppl. II, C402.
This can be interpreted as the proportion of variance in
the observations accounted for by the model forecasts.
This criterion is used for the evaluation of the forecast Moore, R. J., May, B. C., Jones, D. A. & Black, K. B. (1994).
hyetograph and hydrograph.
Local calibration of weather radar over London. In M. E.
Almeida-Teixeira, R. Fantechni, R. Moore & V. M. Silva
(eds.) Advances in Radar Hydrology, European
Commission Report EHR 14334 EN, 186–195.
Ogura, Y. & Takahashi, T. (1971). Numerical simulation of
the life cycle of a thunderstorm. Mon. Wea. Rev., 99:
895–911.
Oki, T., Musiake, K. & Koike, T. (1991). Spatial rainfall dis-
tribution at a storm event in mountainous regions, esti-
mated by orography and wind direction. Water Resour.
Res., 27: 359–369.
References
Alpert, P. (1986). Mesoscale indexing of the distribution of
orographic precipitation over high mountains. J. Climate
Appl. Meteorol., 25: 532–545.
Alpert, P. & Shafir, H. (1989). Mesoγ-scale distribution of
orographic precipitation: Numerical study and comparison
with precipitation derived from radar measurement. J.
Appl. Meteorol., 28: 1105–1117.
Sinclair, M. R. (1994). A diagnostic model for estimating oro-
graphic precipitation. J. Appl. Meteorol., 33: 1163–1175.
Bell, R. S. (1978). The forecasting of orographically enhanced
3
05