Ambipolar Diffusion and Dynamo Requirements

Art provided a list of variables needed for ionospheric (and electrodynamic?) calculations to be added to the WACCM-X h5.model. Here is a list of the variables and ideas of where to access these variables in the current model:

• Magnetic latitude and longitude: Apex coordinates used in WACCM. Variables called 'alatm' and 'alonm' in mo_apex.F90 module called in module iondrag.F90, Need to add these to the 'pbuf' physics buffer with pbuf_add subroutine.

• Magnetic field components: Calculated in mo_apex.F90 module called in module iondrag.F90. B _y is variable 'bnorth', B _x is variable 'beast', B _z is variable 'bdown', and summation of 3 components is variable 'bmag'. These are 2D variables but need to be expanded to 3D variables to include the vertical dimension

• ExB/B ² components(v): Calculated in exbdrift.F90 module which is called in iondrag.F90 module. v _x is variable 'ue', v _y is variable 've' and v _z is variable 'we'. These are 2D variables and need to be expanded in the vertical to 3D. v _z needs to be converted to interface levels('wei'). 2D 'ue', 've', and 'we' are stored in the 'pbuf' physics buffer.

• Temperature and winds: T, U, V are in 'state' physics structure at midpoint vertical levels. W is in 'state' physics structure at midpoints named 'omega'. Need to convert to interface levels and add to 'state' structure ('omegai' or 'wi'?) and update each time step.

• n(O),n(O ₂),n(N ₂),n(H),n(He): All but He in 'state' structure 'q' array. He not critical now so revisit later.

• Chapman function radiation attenuation near terminator?

• Scale height: Needs to be calculated and added to 'state' structure and update each time step

• Gravity: Would prefer altitude dependent but revisit later

• Geopotential altitude: In 'state' structure as 'zm' at midpoint and 'zi' at interface levels

• Zenith angle: Cosine of zenith angle calculated in zenith.F90 module

• Masses of molecules: 'const_mw' array in constituents.F90 module accessed through 'cnst_get_ind' subroutine

• Masses of ions: Found ions of O,O ₂,N,N ₂,NO but not H and He in 'q' array of 'state' structure. Need to add H (and He later).

• Ion production rates: Need to locate where production rates calculated, maybe near end of chemistry? To solve for transport due to ambipolar diffusion, we will use implicit solver. The production/loss terms form the RHS of the solver, which should help assuring numerical stability.

• Solar spectrum and auroral spectra: Solar spectrum need to look at photochemistry to see where used. Probably in mo_aurora.F90 module accessed in iondrag.F90 module. (We will get a total production/loss rate, and don't need to get individually ion/electron production from solar photoionization or auroral production.)

• Heating rates: Look in mo_waccm_hrates.F90 module?

• Conduction coefficients at interface levels for n(O),n(N ₂),n(O ₂),n(H),n(He),T,[T _e,n _i, n _e]: Molecular conductivity in 'state' structure calculated based on O ₂, N ₂, and O. (The heating rates and thermal conduction are needed for calculating the electron/ion temperature, which will be done before the transport.

• Values at previous time of n(O ⁺),n(H ⁺),n(He ⁺),T _i,T _e,[n(O ₂⁺),n(NO ⁺),...]: 'state' structure contains previous time step values until updated. Need to find where each updated but H ⁺ and He ⁺ are not present so H ⁺ needs to be added (He later).

• Boltzmann constant: Use shr_const_mod variable is shr_const_kboltz

• Electron number density: In 'state' structure variable 'q' state%q(i,k,id_elec). Units are cm^-3^?

• Dip angle: calculated in iondrag_calc and needs to be put in physics buffer 'pbuf'

• Neutral temperature: In 'state' structure variable 't' midpoints. Need to convert to interface like in diffusion_solver.F90 module

• Electron temperature: set to neutral temperature at first step and calculate and update in energy routine

• O,O ₂,N ₂ (and ions of these) number density: Convert from 'state' structure 'q' variable mmrs using 'invariants'?, mmr2vmr?, similar to what is done in mo_waccm_photo.F90 module

• Pressure: In 'state' structure 'pmid' variable in Pascals

• Constituent dependent gas constant: In 'state' structure 'rairv' variable

Notes:

Where interface variables are needed and not currently available need to calculate and add 'i' to midpoint name

Dip angle is calculated from magnetic components (already done in iondrag_calc)

iondrag_calc subroutine from iondrag.F90 module called near end of 'tphysac' routine (tphysac.F90 module) which is called from 'physrun2' routine (physpkg.F90 module) for each chunk. 'physrun2' is called from 'cam_run2' routine (cam_comp.F90 module)