Assignment 3

 

Achieve 80% to pass

 

  1. Suppose a slow computer system can only use 1 wall clock hour of execution time to simulate 2 hours in the Numerical Weather Prediction model, and due to the collection and preparation of Initial Condition data, the NWP simulation can start 6 hours after the real-time. When would a forecast with T+132 hours valid time (5 days' forecast with a model spin-up of 12 hours) be available? (T is the date-time of the initial conditions)
    1. T+6hour
    2. T+60hour
    3. T+63hour
    4. T+72hour

 

  1. Suppose the computer system in the above question is fast and can use 1 wall clock hour of execution time to simulate 20 hours in the Numerical Weather Prediction model. What is the answer to the same question?
    1. T+6.6hour
    2. T+12.6hour
    3. T+18.6hour
    4. T+60.6hour

 

  1. Suppose a tropical cyclone approaching Luzon island is expected to have an eyewall with a radius of 56km from the storm centre. You would like to use a numerical model to predict and study eyewall replacement and evolution of the TC for the coming 5 days. Select all statements below that sound reasonable:
    1. I can use a regional model and report results on the 4th-5th days with the same confidence as the 1st-3rd days.
    2. It is okay to use 48km horizontal resolution around tropical cyclones.
    3. I probably need to use 16km or finer resolution around the tropical cyclone.
    4. I shall try several combinations of reasonable choices of physics parameterization schemes and the parameters therein, and pay attention to the variabilities among the results.

 

  1. Somebody uses a regional model to simulate a historic tropical cyclone for 3 days. He uses a single domain with the highest resolution he can afford, which is an order of magnitude higher than the driving data for the lateral boundary condition. The domain he configured tightly contains the track of the TC and its spiral rainbands. What comments below sound reasonable to him?
    1. Using a regional model is okay. By providing suitable lateral boundary condition data, he does not need to spend computational resources to simulate the whole globe.
    2. The domain is wide enough because the important feature to simulate, the rainbands, are already contained in the domain.
    3. The domain should be made wider because the lateral boundary should keep a distance from the simulated features, to minimize the impact of undesirable effects caused by the lateral boundaries.
    4. He is advised to use domain nesting, because the outer domain(s) can cover a wider area and use a more consistent model (same as the innermost domain, or with parameterization schemes not applicable to high resolution disabled) than the driving data. 

 

  1. Which of the following statements is incorrect?
    1. If the Limited-area mesh option in CPAS is used, the corresponding real simulation becomes a boundary value problem.
    2. To solve a boundary value problem of NWP’s partial differential equations, only boundary condition data is required, and initial condition data is not necessary.
    3. The boundary condition can be seen as a constraint to the partial differential equations.
    4. The physically-based governing equations for NWP respects the conservation of mass, momentum and energy. The numerical algorithm for solving the PDEs maintains the conservation as far as possible.

 

  1. What are the sources of state imbalance when applying atmospheric state data from an external source to a NWP model? Select all that apply:
    1. Some categories of hydrometers (moisture in different categories, e.g. cloud water) may be absent in the external source.
    2. The external source may use a different vertical coordinate system with different vertical spacing. The interpolation method may not handle or represent the vertical structure of the atmospheric variables well.
    3. The terrain heights represented by the external data may have big differences from those represented by the current NWP model
    4. The land use/landcover data of the external source is different from those of the current NWP model.
    5. Data assimilation of observation data produces a statistical estimate of the atmospheric spatial field, which may not be consistent with the dynamical system the NWP model is representing.

 

  1. Which of the following statements are correct, assuming the NWP model and software are correctly implemented?
    1. Specifying the NWP software to output more diagnostic variables will alter simulation results.
    2. A NWP model with more prognostic variables is always more accurate than those with less.
    3. For each prognostic variable at each grid point, the NWP needs to effectively calculate the tendency of the prognostic variable as the difference to apply from the current timestep to the next timestep.
    4. All diagnostic variables are 3-dimensional. 

 

  1. Select all statements below that reasonably describe common NWP models and/or the CPAS platform:
    1. The arrangement of physical processes into different physics parameterization schemes is strict and exclusive. There is no overlapping in representations of sub-grid physical processes between schemes.
    2. The development of physics parameterization schemes have been ever evolving. Users should keep themselves aware of choices and new developments, and updates of the schemes across versions.
    3. The choice of schemes may affect the availability of output prognostic and diagnostic variables.
    4. One flexibility of domain nesting support in NWP models (e.g. WRF) is that the parameterization scheme to be used in each domain can be specified separately. For example, some parameterization schemes can be explicitly switched off for high resolution domains.

 

  1. Sort the requirement of computational resources by common implementations of numerical approaches for turbulence, among Direct numerical simulation (DNS), Large eddy simulation (LES), and Reynolds-averaged Navier-Stokes equation (RANS):
    1. RANS > LES > DNS
    2. DNS > RANS > LES
    3. LES > RANS > DNS
    4. DNS > LES > RANS
    5. None of above

 

  1. Which statement(s) below for the Planetary Boundary Layer is/are correct:
    1. In daytime, turbulence is usually driven by buoyancy; in nighttime, the portion of turbulence caused by vertical wind shear increases.
    2. The PBL height usually increases after sunset.
    3. Turbulence kinetic energy becomes zero after sunset.
    4. A stable boundary layer usually develops in the daytime

 

Previous   Answer