Low (web) Quality - BALTEX
Low (web) Quality - BALTEX
Low (web) Quality - BALTEX
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
151<br />
and 30km). Domains D1 (90 and 60-km) and D7 (90-km)<br />
were considered for only coarser grid distance due to<br />
computational constraints. Domains D4 and D5 which have<br />
same longitudinal extents but different latitudinal extents<br />
were re-run to verify and validate results at finer grid<br />
distance. As mentioned above, the larger scales simulated in<br />
model dynamics will depend not only on the size of the<br />
domain but also on the geographical coverage. The seven<br />
domains were thus chosen to provide an ensemble of largescale<br />
dynamics and terrain effects as affected by different<br />
domains. For each of the domains mentioned above, the<br />
model was integrated for 5 days starting from five different<br />
initial conditions 6-hours apart beginning from 00UTC, 24 th<br />
July 2005. Although the event was highly localized in both<br />
space and time, we have considered a spatial window of<br />
2 o x2 o centered on the general event location and a time<br />
window of 24-hour (06UTC, 26 th July to 06UTC, 27 th July<br />
2005) for diagnosis and analysis.<br />
D6<br />
D1<br />
D4<br />
D2<br />
Figure 1. Experimental domains.<br />
D5<br />
D7<br />
D3<br />
5. Benchmark Simulation<br />
Due to large number of simulations involved , high<br />
resolution simulations (10-km) were carried out for only two<br />
domains (D4 and D5) which are comparable in size but vary<br />
in geographical coverage. These simulations and their<br />
comparison with observations are then used as benchmark<br />
for the acceptability of simulations with coarser resolutions.<br />
We then, compared the spatial distribution of 24-hour<br />
accumulated simulated rainfall for these domains at four<br />
horizontal resolutions (10, 30, 60 and 90-km) and found that<br />
a localized intense rainfall event around Mumbai was seen<br />
for all the resolutions and for both the domains. In<br />
particular, both high (10-km) and coarse (30-km) resolution<br />
simulations could capture the distribution reasonably well<br />
especially in terms intensity when compared with satellite<br />
data.<br />
6. Results and Discussion<br />
Comparison of spatial distributions of 24-hour (event<br />
window) accumulated, simulated ensemble rainfall based on<br />
five initial conditions for five domains of 30-km resolution<br />
with satellite data has shown considerable inter-domain<br />
variations in the simulated distribution of rainfall.<br />
Simulations with resolutions of 90 and 60 km also have<br />
shown significant variations in the distribution and intensity<br />
of the simulated rainfall with the size and geographical<br />
coverage of the domain. This inter-domain variability in the<br />
distribution of rainfall is profound at higher grid spacing. It<br />
was also found that the domain D5 (which included<br />
equatorial belt) simulated the observed distribution better<br />
than all the other domains, and for all the three resolutions.<br />
The effects of changes due to longitudinal or latitudinal<br />
extents or both on the 24-hour accumulated area averaged<br />
rainfall (R av ) and the maximum rainfall (R max ) were<br />
analyzed. The results are from ensemble average<br />
simulations with the five leads (initial conditions)<br />
described above. It was shown that changes in the domain<br />
size results in significant changes in R av and R max, it was<br />
true for domains of different resolutions as well. This<br />
variability of R av and R max due to change in domain size<br />
and coverage in terms of Standard Deviation (SD) as<br />
percentage of mean across different domains was in the<br />
range of 30 to 40 % and 34 to 40 % respectively. It was<br />
seen that effects of change in the domain size in general<br />
are larger for lower resolutions. This implies the need for<br />
an optimum model configuration with appropriate domain<br />
size or ensemble of domains and resolution to resolve both<br />
ends of the spectrum of scales. We next examined the<br />
relative role of resolution on R av and R max. Dispersion of<br />
forecasts due to changes in initial conditions provides both<br />
a measure of reliability of the forecasts and a measure of<br />
variation in mesoscale forecasts due to changes in the<br />
large scale conditions (initial fields). In particular, we<br />
expect the response to different initial conditions to be a<br />
strong function of the domain size and geographical<br />
coverage as the large scale fields evolve differently over<br />
different domains. Summary of the relative variability or<br />
sensitivity of R av and R max in terms of SD (as % mean)<br />
with respect to different domains, resolutions and IC’s are<br />
summarized in Table.1. As another measure of relative<br />
sensitivity of the simulations to domain size we have<br />
considered standard deviation (as % of mean) in terms of<br />
24-hour accumulated total rain over event location (R T )<br />
for different domains (with varying number of domains<br />
depending on the resolution). Time evolution of this<br />
standard deviation indicated that variation due to change<br />
of domain is typically 40 % and can be as much as 70 to<br />
80%; in comparison a change of resolution doesn’t change<br />
this dispersion that significantly except at isolated hours<br />
and not more than by a factor of 2.<br />
Table 1. Summary of variability of R av and R max in terms<br />
of SD as % of mean for different domains, resolution and<br />
Initial Conditions.<br />
Rain<br />
References<br />
Model<br />
Domain<br />
Resolution<br />
Initial<br />
Condition<br />
R av 30-40 10-30 11-50<br />
R max 34-40 21-43 6-40<br />
Dudhia, J., PSU/NCAR Mesoscale Modeling System<br />
Tutorial Class Notes and User’s Guide; MM5 Modeling<br />
System Version 3, 2004<br />
Giorgi, F., and L.O. Mearns, Introduction to special<br />
section: Regional climate modeling revisited. J.Geophys.<br />
Res., Vol.104, No. 6, pp. 6335-6352, 1999<br />
Pielke, R.A., Mesoscale Meteorological Modeling.<br />
Academic Press, pp. 673, 2002<br />
Warner, T.T., Peterson, R.A. and Treadon R.E., Tutorial<br />
on lateral boundary condition as a basic and potentially<br />
serious limitation to regional numerical weather<br />
prediction. Bull. of the Amer.Meteor.Soc., No.78, pp.<br />
2599-2617, 1997