WRF-DFI project meeting – 20060718

Hans, Wei, Min, Ju-Won

Lilly room, 2:00-3:30pm, Tuesday 18 Jul 2006

1.Work review

1.1. DFL

Ju-Won made a cycling experiment with DFL, DFL, over a 7-day period from 12UTC 04 May 2006 to 12UTC 11 May 2006. Some results were compared with two reference experiments, CYC (cycling experiment without initialization) and COL (cold start experiment). There is a clear reduction in noise in DFL, compared to CYC and COL. Hans asked for verification scores, both for conventional variables and for precipitation.

The oscillations in moisture and heat flux tendency still exist. There were oscillations in precipitation rate in the first cycle, but not in the last cycle. Wei suggested separating convective precipitation from the total precipitation to see where the oscillations come from.

Wei said one cannot use mu_1 and mu_2 to compute the mu tendency. She suggested changing the calculation for the MU tendency following that for the surface pressure tendency to use mu_2.

Hans found that the time series pattern of domain average divergence is different from pressure tendency at each level in MM5 and asked Min to check it.

Hans said the cutoff time could be tuned, but based on the DFL results he thought the tuning could be wait.

Min made the DFL text. Ju-Won also made a diagnosis text, though it need more text. Hans will work on the documents.

1.2. Backward integration

Wei and Ju-Won made the lateral boundary data every 3hr for backward integration. But the horizontal diffusion problem still exists. Wei will check out that, first. There are two issues: implicit and explicit. Wei made some progress in the implicit diffusion and agreed to give the code to Min for real data tests. With 10 km domain, explicit diffusion is needed. Wei agreed to take a look into the explicit diffusion in backward integration.

2.Work plan for two weeks

All: try to understand the strange oscillations

Hans: test the latest code; review the notes.

Wei: check the horizontal diffusion, MU problem, program to calculate for a single point.

Min: check the horizontal diffusion problem with Wei first, and then test the backward integration.

Ju-Won: fix the MU calculation with Wei; verify the DFL test; print some variables on a single point.

3.AOB

Wei will take vacation starting August 7. We should get all technical related issues solved before her vacation.

Hans stressed the delivery time, the end of August.

We will have the next project meeting on: Tuesday 1 Aug, 2 pm, Lilly room.