03.04.2013 Views

Advanced Research WRF (ARW) Technical Note - MMM - University ...

Advanced Research WRF (ARW) Technical Note - MMM - University ...

Advanced Research WRF (ARW) Technical Note - MMM - University ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

In the current version of the <strong>ARW</strong>, the advection of vector quantities (momentum) and scalars<br />

is performed using the RK3 time integration as outlined in Fig. 3.1. The spatial discretization<br />

used in this approach is outlined in the next section. For many applications it is desirable to<br />

use positive definite or monotonic advection schemes for scalar transport. In the next major<br />

release of the <strong>ARW</strong> we will be including a forward-in-time scheme for scalar transport that has<br />

positive definite and monotonic options. We describe that scheme in the section following the<br />

description of the RK3 advection.<br />

RK3 Advection<br />

2 nd through 6 th order accurate spatial discretizations of the flux divergence are available in the<br />

<strong>ARW</strong> for momentum, scalars and geopotential using the RK3 time-integration scheme (scalar<br />

advection option 1, step 7 in the time-split integration sequence in Fig. 3.1). The discrete<br />

operators can be illustrated by considering the flux divergence equation for a scalar q in its<br />

discrete form:<br />

R t∗<br />

qadv = −m2 [δx(Uq xadv ) + δy(V q yadv )] − mδη(Ωq ηadv ). (3.39)<br />

As in the pressure gradient discretization, the discrete operator is defined as<br />

δx(Uq xadv −1<br />

) = ∆x (Uq xadv<br />

)i+1/2 − (Uq xadv<br />

<br />

)i−1/2 . (3.40)<br />

The different order advection schemes correspond to different definitions for the operator q xadv.<br />

The even order operators (2 nd , 4 th , and 6 th ) are<br />

2 nd order: (q xadv )i−1/2 = 1<br />

2 (qi + qi−1)<br />

4 th order: (q xadv )i−1/2 = 7<br />

12 (qi + qi−1) − 1<br />

12 (qi+1 + qi−2)<br />

6 th order: (q xadv )i−1/2 = 37<br />

60 (qi + qi−1) − 2<br />

15 (qi+1 + qi−2) + 1<br />

60 (qi+2 + qi−3),<br />

and the odd order operators (3 rd and 5 th ) are<br />

3 rd order: (q xadv )i−1/2 = (q xadv ) 4 th<br />

i−1/2<br />

+ sign(U) 1<br />

12<br />

(qi+1 − qi−2) − 3(qi − qi−1) <br />

5 th order: (q xadv<br />

)i−1/2 = (q xadv 6<br />

) th<br />

i−1/2<br />

− sign(U) 1 <br />

(qi+2 − qi−3) − 5(qi+1 − qi−2) + 10(qi − qi−1)<br />

60<br />

.<br />

The even-order advection operators are spatially centered and thus contain no implicit diffusion<br />

outside of the diffusion inherent in the RK3 time integration. The odd-order schemes are<br />

upwind-biased, and the spatial discretization is inherently diffusive. As is evident in their formulation,<br />

the odd-order schemes are comprised of the next higher (even) order centered scheme<br />

plus an upwind term that, for a constant transport mass flux, is a diffusion term of that next<br />

higher (even) order with a hyper-viscosity proportional to the Courant number (Cr). Further<br />

details concerning RK3 advection can be found in Wicker and Skamarock (2002)<br />

19

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!