289
Precipitation Estimate Using NEXRAD Ground-Based Radar Images
mm) is much higher than that observed at P44 (1609 mm). For each precipitation
event, given the availability of rain gauge network with high density, the coefficient
of variation (CV) is calculated using observations from rain gauges instead of using
precipitation estimates from NEXRAD grids. To examine the effect of the subgrid
heterogeneity of precipitation distribution on NEXRAD performance, we followed
Habib et al. (2009) and classified precipitation events into three categories: uniform
(CV < 0.2), medially variable (0.2 ≤ CV < 0.5), and highly variable (0.5 ≤ CV). The
evaluation coefficients were calculated to illustrate the performance of NEXRAD
data under different precipitation variability conditions.
The evaluation coefficients were calculated using precipitation events with uniform, medially variable, and highly variable distribution, respectively (Tables 12.2
and 12.3). For the I grid, NEXRAD performance decreases as precipitation variability increases. For example, ranges of D rain are 88%–96%, 71%–72%, and 35%–
59%, respectively, under uniform, medially variable, and highly variable conditions.
Values of R 2 decrease from 0.53–0.59 for uniform events to 0.21–0.31 for medially
TABLE 12.3
Validation of NEXRAD for Rain Events with Different Spatial Variability for
the G Grid
P44
P46
P48
P49
Gauge
Mean
NEXRAD
Uniform
Accumulated
precipitation
(mm)
713.98
799.83
773.16
730.25
754.31
730.37
D rain
95%
97%
97%
97%
97%
D no-rain
98%
98%
98%
98%
98%
EB
2%
–9%
–6%
0%
–3%
R 2
0.83
0.90
0.91
0.90
0.90
EE
0.82
0.90
0.91
0.90
0.89
Medially
variable
Accumulated
precipitation
(mm)
731.98
1396.26
1307.51
1218.42
1163.54
1113.12
D rain
89%
87%
91%
89%
86%
D no-rain
38%
100%
71%
67%
–
EB
52%
–20%
–15%
–9%
–4%
R 2
0.89
0.89
0.93
0.97
0.95
EE
0.45
0.86
0.89
0.94
0.95
Highly
variable
Accumulated
precipitation
(mm)
153.78
269.14
226.50
162.21
202.91
276.14
D rain
58%
73%
82%
83%
74%
D no-rain
65%
78%
77%
77%
82%
EB
80%
3%
22%
70%
36%
R 2
0.18
0.27
0.41
0.76
0.56
EE
–2.85
0.16
0.29
0.47
0.09
Precipitation Estimate Using NEXRAD Ground-Based Radar Images
mm) is much higher than that observed at P44 (1609 mm). For each precipitation
event, given the availability of rain gauge network with high density, the coefficient
of variation (CV) is calculated using observations from rain gauges instead of using
precipitation estimates from NEXRAD grids. To examine the effect of the subgrid
heterogeneity of precipitation distribution on NEXRAD performance, we followed
Habib et al. (2009) and classified precipitation events into three categories: uniform
(CV < 0.2), medially variable (0.2 ≤ CV < 0.5), and highly variable (0.5 ≤ CV). The
evaluation coefficients were calculated to illustrate the performance of NEXRAD
data under different precipitation variability conditions.
The evaluation coefficients were calculated using precipitation events with uniform, medially variable, and highly variable distribution, respectively (Tables 12.2
and 12.3). For the I grid, NEXRAD performance decreases as precipitation variability increases. For example, ranges of D rain are 88%–96%, 71%–72%, and 35%–
59%, respectively, under uniform, medially variable, and highly variable conditions.
Values of R 2 decrease from 0.53–0.59 for uniform events to 0.21–0.31 for medially
TABLE 12.3
Validation of NEXRAD for Rain Events with Different Spatial Variability for
the G Grid
P44
P46
P48
P49
Gauge
Mean
NEXRAD
Uniform
Accumulated
precipitation
(mm)
713.98
799.83
773.16
730.25
754.31
730.37
D rain
95%
97%
97%
97%
97%
D no-rain
98%
98%
98%
98%
98%
EB
2%
–9%
–6%
0%
–3%
R 2
0.83
0.90
0.91
0.90
0.90
EE
0.82
0.90
0.91
0.90
0.89
Medially
variable
Accumulated
precipitation
(mm)
731.98
1396.26
1307.51
1218.42
1163.54
1113.12
D rain
89%
87%
91%
89%
86%
D no-rain
38%
100%
71%
67%
–
EB
52%
–20%
–15%
–9%
–4%
R 2
0.89
0.89
0.93
0.97
0.95
EE
0.45
0.86
0.89
0.94
0.95
Highly
variable
Accumulated
precipitation
(mm)
153.78
269.14
226.50
162.21
202.91
276.14
D rain
58%
73%
82%
83%
74%
D no-rain
65%
78%
77%
77%
82%
EB
80%
3%
22%
70%
36%
R 2
0.18
0.27
0.41
0.76
0.56
EE
–2.85
0.16
0.29
0.47
0.09