todo

# TODO List for iers
--------------------

## Fix Mock Tests
-----------------
test/should_not_compile does not work!

## ICGEM
---------

  * when parsing ICGEM file headers, only a number of error identifiers are 
  valid (e.g. 'calibrated_and_formal', 'calibrated', etc). Add 'empirical'.

  * add custom keyword/value pais so that we can store information like: 
  time_period which are non-standard.

### BUG
-------

  skipws should also handling a leading '+' sign in floating point. 
  std::from_chars does not allow for a leading '+' sign.

## Solid earth tides
---------------------

  Displacement for SET differs from dehanttideinel respective results below 
  1mm but i am not hapy with that! Basically, the computation block that causes 
  the discrepancies is the diurnal corections for Step-2. 
  If i use the full set of diurnal waves (see Tables7Extended in 
  iers2010_displacement_step2.cpp) then results show an order of magnitude 
  better compliance w.r.t. DEHANTTIDEINEL. This extended tables includes 31 
  diurnal waves extending the original 11 waves included in Table 7a of the 
  IERS standards.

  See the unit test code test/unit_tests/solid_earth_tide_displacement.cpp.
  The input (validation) data used therein are computed using dehanttideinel

## Atmospheric Tides
--------------------

  To construct an atmospheric tide instance, one of the following two ways can 
  be used:
  1. Input tidal wave data from AOD1B files (i.e. for each wave, e.g. 
    using AOD1B_ATM_R2_06.asc, AOD1B_ATM_S1_06.asc, etc). This is checked for 
    the AOD1B RL06 series (via the COST-G benchmark) but a new version is now 
    available. THe thing is, that this way we have to use a conventional π 
    factor when computing each tidal contribution which is a bit messed up. 
    There is no clear way of deriving this factor (see 
    src/lib/atmospheric_tides/atm_tides_geopotential.cpp).

  2. Input tidal wave data from .gfc file, the GROOPS way. This presumably 
    works, but is untested!