Impact of Size and Coverage of the Model Domain on Simulation of Mesoscale Events
IR@C-MMACS: CSIR-Centre for Mathematical Modelling and Computer Simulation, Bangalore
View Archive Info| Field | Value | |
| Title |
Impact of Size and Coverage of the Model Domain on Simulation of
Mesoscale Events
|
|
| Creator |
S, Himesh
Goswami, P |
|
| Subject |
Climate Modelling
|
|
| Description |
The most commonly used current strategy for mesoscale simulation and forecasting is to use Limited Area Models
(LAM), three dimensional models with artificial lateral boundary conditions. While these models can support
high horizontal resolution necessary to resolve surface inhomogenities and convective dynamics at scales not
yet feasible in General Circulation Models (GCM), a necessary price is the artificial lateral boundaries. The limited
domain introduces several aspects to a mesoscale simulation. In particular, there is no unique choice of the
extent and geographical coverage of the domain. Besides, the large-scale forcing at the lateral boundaries may
not always be dynamically consistent with the inner (mesoscale) fields. The question of the lateral boundary or
the domain size is inherently related to model dynamics and physics. While resolution determines the smallest
resolved scale, the domain size restricts the largest (horizontal) scale resolved. Thus, domain size and resolution
together determine the spectrum of resolved scale and nature of scale interaction in the model dynamics.
The geographical coverage (necessarily covering the event location) of the mesoscale simulation domain also
plays a significant role by selectively including or excluding processes like orographic lifting and equatorial
waves. In this work we present a detailed investigation of the effect of size and coverage of the domain on the
quality of simulation of a mesoscale event. We have used the mesoscale model MM5 to investigate the extreme
rainfall event that occurred over Mumbai over the west coast of India during 26-27 July, 2005. For each of the
five chosen domains with different spatial extent and geographical coverage, simulations were carried out with
five initial conditions (leads) for a given resolution with identical physics options. These simulations were repeated
for different resolutions (90, 60 and 30km), resulting in a total of 75 simulation experiments. In addition,
two larger (one nearly global) domains were considered for coarser resolution (90-60km). Our results show that
along with resolution, the geographical coverage and the size of the domain also play critical roles in the simulation
of a mesoscale event. The best simulation is found to be for domains that are not the largest but that cover
significant part of the equatorial ocean. While this conclusion is likely to change based on event location, our
study shows that the choice of the meso-scale domain is a non-trivial but critical input for improved mesoscale
simulation and forecasting, and needs to be determined through a comprehensive calibration experiment.
|
|
| Publisher |
CSIR Centre for Mathematical Modelling and Computer Simulation
|
|
| Date |
2007
|
|
| Type |
Monograph
NonPeerReviewed |
|
| Format |
application/pdf
|
|
| Identifier |
http://cir.cmmacs.ernet.in/268/1/rrcm%2D0701.pdf
S, Himesh and Goswami, P (2007) Impact of Size and Coverage of the Model Domain on Simulation of Mesoscale Events. Technical Report. CSIR Centre for Mathematical Modelling and Computer Simulation , C-MMACS,Bangalore 560037,India. (Unpublished) |
|
| Relation |
http://cir.cmmacs.ernet.in/268/
|
|