Changes between Version 7 and Version 8 of AvailableOptionsforInitPDAF


Ignore:
Timestamp:
Dec 22, 2014, 4:31:53 PM (9 years ago)
Author:
lnerger
Comment:

--

Legend:

Unmodified
Added
Removed
Modified

  • AvailableOptionsforInitPDAF

    v7 v8  
    1111
    1212{{{
    13         Available options:
    14 
    15         Sub-types (Parameter subtype)
    16           0: Evolve unit modes with finite difference approx. of TLM
    17           1: like 0 with modes scaled by eigenvalues, unit U
    18           2: Fixed basis vectors; variable U matrix
    19           3: Fixed covariance matrix (V and U kept constant)
    20           5: Offline mode
    21 
    22         Integer parameters (Array param_int)
    23           param_int(1): Dimension of state vector (>0), required
    24           param_int(2): Ensemble size (>0), required
    25           param_int(3): Interval for re-diagonalization of P (>0); optional: default 1
    26           param_int(4): 1 for incremental updating, 0 else; optional: default 0
    27 
    28         Floating point parameters (Array param_real)
    29           param_real(1): Forgetting factor (usually >0 and <=1), required
    30           param_real(2): epsilon for finite-difference approx. of TLM, required
    31 
    32         Further parameters
    33           n_modeltasks: Number of parallel model integration tasks
    34               >=1 for subtypes 0 and 1; not larger than total number of processors
    35               =1 required for subtypes 2 and 3
    36           screen: Control verbosity of PDAF
    37               0: no outputs
    38               1: basic output (default)
    39               2: 1 plus timing output
    40               3: 2 plus debug output
    41 
    42         +++++++++ End of option overview for the SEEK filter ++++++++++
     13PDAF     Available options for SEEK:
     14PDAF     --- Sub-types (Parameter subtype) ---
     15PDAF       0: Evolve unit modes with finite difference approx. of TLM
     16PDAF       1: like 0 with modes scaled by eigenvalues, unit U
     17PDAF       2: Fixed basis vectors; variable U matrix
     18PDAF       3: Fixed covariance matrix (V and U kept constant)
     19PDAF       5: Offline mode
     20PDAF     --- Integer parameters (Array param_int) ---
     21PDAF       param_int(1): Dimension of state vector (>0), required
     22PDAF       param_int(2): Ensemble size (>0), required
     23PDAF       param_int(3): Interval for re-diagonalization of P (>0); optional: default 1
     24PDAF       param_int(4): 1 for incremental updating, 0 else; optional: default 0
     25PDAF     --- Floating point parameters (Array param_real) ---
     26PDAF       param_real(1): Forgetting factor (usually >0 and <=1), required
     27PDAF       param_real(2): epsilon for finite-difference approx. of TLM, required
     28PDAF     --- Further parameters ---
     29PDAF       n_modeltasks: Number of parallel model integration tasks
     30PDAF           >=1 for subtypes 0 and 1; not larger than total number of processors
     31PDAF           =1 required for subtypes 2 and 3
     32PDAF       screen: Control verbosity of PDAF
     33PDAF           0: no outputs
     34PDAF           1: basic output (default)
     35PDAF           2: 1 plus timing output
     36PDAF           3: 2 plus debug output
     37PDAF     +++++++++ End of option overview for the SEEK filter ++++++++++
    4338}}}
    4439
    … …  
    4742
    4843{{{
    49         Available options:
    50 
    51         Sub-types (Parameter subtype)
    52           0: full ensemble integration; left-sided application of T
    53           1: full ensemble integration; right-sided application of T
    54           2: Fixed error space basis
    55           3: Fixed state covariance matrix
    56           4: Implementation with explicit ensemble transformation
    57           5: Offline mode
    58 
    59         Integer parameters (Array param_int)
    60           param_int(1): Dimension of state vector (>0), required
    61           param_int(2): Ensemble size (>0), required
    62           param_int(3): not used
    63           param_int(4): 1 for incremental updating, 0 else; optional, default: 0
    64           param_int(5): Type of forgetting factor; optional, default: 0
    65               0: fixed forgetting factor
    66               1: adaptive forgetting factor (experimental)
    67           param_int(6): Type of ensemble transformation matrix; optional, default: 0
    68               0: deterministic omega
    69               1: random orthonormal omega orthogonal to (1,...,1)^T
    70               2: use product of 0 with random orthonomal matrix with eigenvector (1,...,1)^T
    71                  (experimental; for random transformations, 1 is recommended)
    72           param_int(7): Type of transformation matrix square root; optional, default: 0
    73               (Only relevant for subtype=4)
    74               0: symmetric square root
    75               1: Cholesky decomposition
    76 
    77         Floating point parameters (Array param_real)
    78           param_real(1): Forgetting factor (usually >0 and <=1), required
    79 
    80         Further parameters
    81           n_modeltasks: Number of parallel model integration tasks
    82               >=1 for subtypes 0 and 1; not larger than total number of processors
    83               =1 required for subtypes 2 and 3
    84           screen: Control verbosity of PDAF
    85               0: no outputs
    86               1: basic output (default)
    87               2: 1 plus timing output
    88               3: 2 plus debug output
    89 
    90         Internal parameter (defined inside PDAF)
    91           Nm1vsN: Normalization of covariance matrix; default: 1
    92               0: normalization with 1/(Ensemble size)
    93                  (original SEIK, mainly for compatibility with older studies)
    94               1: normalization with 1/(Ensemble size - 1)
    95                  (sample covariance matrix consistent with other EnKFs)
    96 
    97         +++++++++ End of option overview for the SEIK filter ++++++++++
     44PDAF     Available options for SEIK:
     45PDAF     --- Sub-types (Parameter subtype) ---
     46PDAF       0: full ensemble integration; left-sided application of T
     47PDAF       1: full ensemble integration; right-sided application of T
     48PDAF       2: Fixed error space basis
     49PDAF       3: Fixed state covariance matrix
     50PDAF       4: Implementation with explicit ensemble transformation
     51PDAF       5: Offline mode
     52PDAF     --- Integer parameters (Array param_int) ---
     53PDAF       param_int(1): Dimension of state vector (>0), required
     54PDAF       param_int(2): Ensemble size (>0), required
     55PDAF       param_int(3): not used
     56PDAF       param_int(4): 1 for incremental updating, 0 else; optional, default: 0
     57PDAF       param_int(5): Type of forgetting factor; optional, default: 0
     58PDAF           0: fixed forgetting factor
     59PDAF           1: adaptive forgetting factor (experimental)
     60PDAF       param_int(6): Type of ensemble transformation matrix; optional, default: 0
     61PDAF           0: deterministic omega
     62PDAF           1: random orthonormal omega orthogonal to (1,...,1)^T
     63PDAF           2: use product of 0 with random orthonomal matrix with eigenvector (1,...,1)^T
     64PDAF              (experimental; for random transformations, 1 is recommended)
     65PDAF       param_int(7): Type of transformation matrix square root; optional, default: 0
     66PDAF           (Only relevant for subtype/=3)
     67PDAF           0: symmetric square root
     68PDAF           1: Cholesky decomposition
     69PDAF     --- Floating point parameters (Array param_real) ---
     70PDAF       param_real(1): Forgetting factor (usually >0 and <=1), required
     71PDAF     --- Further parameters ---
     72PDAF       n_modeltasks: Number of parallel model integration tasks
     73PDAF           >=1 for subtypes 0 and 1; not larger than total number of processors
     74PDAF           =1 required for subtypes 2 and 3
     75PDAF       screen: Control verbosity of PDAF
     76PDAF           0: no outputs
     77PDAF           1: basic output (default)
     78PDAF           2: 1 plus timing output
     79PDAF           3: 2 plus debug output
     80PDAF     --- Internal parameter (defined inside PDAF) ---
     81PDAF       Nm1vsN: Normalization of covariance matrix; default: 1
     82PDAF           0: normalization with 1/(Ensemble size)
     83PDAF              (original SEIK, mainly for compatibility with older studies)
     84PDAF           1: normalization with 1/(Ensemble size - 1)
     85PDAF           (sample covariance matrix consistent with other EnKFs)
     86PDAF     +++++++++ End of option overview for the SEIK filter ++++++++++
    9887}}}
    9988
    … …  
    10291
    10392{{{
    104         Available options:
    105 
    106         Sub-types (Parameter subtype)
    107           0: Full ensemble integration; analysis for 2*dim_obs>dim_ens
    108           1: Full ensemble integration; analysis for 2*dim_obs<=dim_ens
    109           5: Offline mode
    110 
    111         Integer parameters (Array param_int)
    112           param_int(1): Dimension of state vector (>0), required
    113           param_int(2): Ensemble size (>0), required
    114           param_int(3): maximum rank for inversion of HPH^T, optional, default=0
    115               (for =0, HPH is inverted by solving the representer equation)
    116               (if set to >=ensemble size, it is reset to ensemble size - 1)
    117           param_int(4): not used
    118           param_int(5): Size of lag for smoothing
    119 
    120         Floating point parameters (Array param_real)
    121           param_real(1): Forgetting factor (usually >0 and <=1), required
    122 
    123         Further parameters
    124           n_modeltasks: Number of parallel model integration tasks
    125               (>=1; not larger than total number of processors)
    126           screen: Control verbosity of PDAF
    127               0: no outputs
    128               1: basic output (default)
    129               2: 1 plus timing output
    130               3: 2 plus debug output
    131 
    132         +++++++++ End of option overview for the EnKF ++++++++++
     93PDAF     Available options for EnKF:
     94PDAF     --- Sub-types (Parameter subtype) ---
     95PDAF       0: Full ensemble integration; analysis for 2*dim_obs>dim_ens
     96PDAF       1: Full ensemble integration; analysis for 2*dim_obs<=dim_ens
     97PDAF       5: Offline mode
     98PDAF     --- Integer parameters (Array param_int) ---
     99PDAF       param_int(1): Dimension of state vector (>0), required
     100PDAF       param_int(2): Ensemble size (>0), required
     101PDAF       param_int(3): maximum rank for inversion of HPH^T, optional, default=0
     102PDAF           (for =0, HPH is inverted by solving the representer equation)
     103PDAF           (if set to >=ensemble size, it is reset to ensemble size - 1)
     104PDAF       param_int(4): not used
     105PDAF       param_int(5): Size of lag for smoothing
     106PDAF     --- Floating point parameters (Array param_real) ---
     107PDAF       param_real(1): Forgetting factor (usually >0 and <=1), required
     108PDAF     --- Further parameters ---
     109PDAF       n_modeltasks: Number of parallel model integration tasks
     110PDAF           (>=1; not larger than total number of processors)
     111PDAF       screen: Control verbosity of PDAF
     112PDAF           0: no outputs
     113PDAF           1: basic output (default)
     114PDAF           2: 1 plus timing output
     115PDAF           3: 2 plus debug output
     116PDAF     +++++++++ End of option overview for the EnKF ++++++++++
    133117}}}
    134118
    … …  
    137121
    138122{{{
    139         Available options:
    140 
    141         Sub-types (Parameter subtype)
    142           0: full ensemble integration; left-sided application of T
    143           2: Fixed error space basis
    144           3: Fixed state covariance matrix
    145           5: Offline mode
    146 
    147         Integer parameters (Array param_int)
    148           param_int(1): Dimension of state vector (>0), required
    149           param_int(2): Ensemble size (>0), required
    150           param_int(3): not used
    151           param_int(4): 1 for incremental updating, 0 else; optional, default: 0
    152           param_int(5): Type of forgetting factor; optional, default: 0
    153               0: fixed forgetting factor
    154               1: adaptive forgetting factor for full domain (experimental)
    155               2: locally adaptive forgetting factor (experimental)
    156           param_int(6): Type of ensemble transformation matrix; optional, default: 0
    157               0: deterministic omega
    158               1: random orthonormal omega orthogonal to (1,...,1)^T
    159               2: use product of 0 with random orthonomal matrix with eigenvector (1,...,1)^T
    160                  (experimental; for random transformations, 1 is recommended)
    161           param_int(7): Type of transformation matrix square root; optional, default: 0
    162               (Only relevant for subtype=4)
    163               0: symmetric square root
    164               1: Cholesky decomposition
    165 
    166         Floating point parameters (Array param_real)
    167           param_real(1): Forgetting factor (usually >0 and <=1), required
    168 
    169         Further parameters
    170           n_modeltasks: Number of parallel model integration tasks
    171               >=1 for subtypes 0 and 1; not larger than total number of processors
    172               =1 required for subtypes 2 and 3
    173           screen: Control verbosity of PDAF
    174               0: no outputs
    175               1: basic output (default)
    176               2: 1 plus timing output
    177               3: 2 plus debug output
    178 
    179         Internal parameter (defined inside PDAF)
    180           Nm1vsN: Normalization of covariance matrix; default: 1
    181               0: normalization with 1/(Ensemble size)
    182                  (original SEIK, mainly for compatibility with older studies)
    183               1: normalization with 1/(Ensemble size - 1)
    184                  (sample covariance matrix consistent with other EnKFs)
    185 
    186         +++++++++ End of option overview for the LSEIK filter ++++++++++
     123PDAF     Available options for LSEIK:
     124PDAF     --- Sub-types (Parameter subtype) ---
     125PDAF       0: full ensemble integration; left-sided application of T
     126PDAF       2: Fixed error space basis
     127PDAF       3: Fixed state covariance matrix
     128PDAF       5: Offline mode
     129PDAF     --- Integer parameters (Array param_int) ---
     130PDAF       param_int(1): Dimension of state vector (>0), required
     131PDAF       param_int(2): Ensemble size (>0), required
     132PDAF       param_int(3): not used
     133PDAF       param_int(4): 1 for incremental updating, 0 else; optional, default: 0
     134PDAF       param_int(5): Type of forgetting factor; optional, default: 0
     135PDAF           0: fixed forgetting factor
     136PDAF           1: adaptive forgetting factor for full domain (experimental)
     137PDAF           2: locally adaptive forgetting factor (experimental)
     138PDAF       param_int(6): Type of ensemble transformation matrix; optional, default: 0
     139PDAF           0: deterministic omega
     140PDAF           1: random orthonormal omega orthogonal to (1,...,1)^T
     141PDAF           2: use product of 0 with random orthonomal matrix with eigenvector (1,...,1)^T
     142PDAF              (experimental; for random transformations, 1 is recommended)
     143PDAF       param_int(7): Type of transformation matrix square root; optional, default: 0
     144PDAF           (Only relevant for subtype/=3)
     145PDAF           0: symmetric square root
     146PDAF           1: Cholesky decomposition
     147PDAF     --- Floating point parameters (Array param_real) ---
     148PDAF       param_real(1): Forgetting factor (usually >0 and <=1), required
     149PDAF     --- Further parameters ---
     150PDAF       n_modeltasks: Number of parallel model integration tasks
     151PDAF           >=1 for subtypes 0 and 1; not larger than total number of processors
     152PDAF           =1 required for subtypes 2 and 3
     153PDAF       screen: Control verbosity of PDAF
     154PDAF           0: no outputs
     155PDAF           1: basic output (default)
     156PDAF           2: 1 plus timing output
     157PDAF           3: 2 plus debug output
     158PDAF     --- Internal parameter (defined inside PDAF) ---
     159PDAF       Nm1vsN: Normalization of covariance matrix; default: 1
     160PDAF           0: normalization with 1/(Ensemble size)
     161PDAF              (original SEIK, mainly for compatibility with older studies)
     162PDAF           1: normalization with 1/(Ensemble size - 1)
     163PDAF              (sample covariance matrix consistent with other EnKFs)
     164PDAF     +++++++++ End of option overview for the LSEIK filter ++++++++++
    187165}}}
    188166
    … …  
    191169
    192170{{{
    193         Available options:
    194 
    195         Sub-types (Parameter subtype)
    196           0: full ensemble integration; apply T-matrix analogously to SEIK
    197           1: full ensemble integration; formulation without T matrix
    198           2: Fixed error space basis; analysis with T-matrix
    199           3: Fixed state covariance matrix; analysis with T-matrix
    200           5: Offline mode; analysis with T-matrix
    201 
    202         Integer parameters (Array param_int)
    203           param_int(1): Dimension of state vector (>0), required
    204           param_int(2): Ensemble size (>0), required
    205           param_int(3): Size of lag for smoothing
    206           param_int(4): not used
    207           param_int(5): Type of forgetting factor; optional, default: 0
    208               0: fixed forgetting factor
    209               1: adaptive forgetting factor (experimental)
    210           param_int(6): Type of ensemble transformation matrix; optional, default: 0
    211               0: deterministic transformation
    212               2: use product of 0 with random orthonomal matrix with eigenvector (1,...,1)^T
    213 
    214         Floating point parameters (Array param_real)
    215           param_real(1): Forgetting factor (usually >0 and <=1), required
    216 
    217         Further parameters
    218           n_modeltasks: Number of parallel model integration tasks
    219               >=1 for subtypes 0 and 1; not larger than total number of processors
    220               =1 required for subtypes 2 and 3
    221           screen: Control verbosity of PDAF
    222               0: no outputs
    223               1: basic output (default)
    224               2: 1 plus timing output
    225               3: 2 plus debug output
    226 
    227         +++++++++ End of option overview for the ETKF ++++++++++
     171PDAF     Available options for ETKF:
     172PDAF     --- Sub-types (Parameter subtype) ---
     173PDAF       0: full ensemble integration; apply T-matrix analogously to SEIK
     174PDAF       1: full ensemble integration; formulation without T matrix
     175PDAF       2: Fixed error space basis; analysis with T-matrix
     176PDAF       3: Fixed state covariance matrix; analysis with T-matrix
     177PDAF       5: Offline mode; analysis with T-matrix
     178PDAF     --- Integer parameters (Array param_int) ---
     179PDAF       param_int(1): Dimension of state vector (>0), required
     180PDAF       param_int(2): Ensemble size (>0), required
     181PDAF       param_int(3): Size of lag for smoothing
     182PDAF       param_int(4): not used
     183PDAF       param_int(5): Type of forgetting factor; optional, default: 0
     184PDAF           0: fixed forgetting factor
     185PDAF           1: adaptive forgetting factor (experimental)
     186PDAF       param_int(6): Type of ensemble transformation matrix; optional, default: 0
     187PDAF           0: deterministic transformation
     188PDAF           2: use product of 0 with random orthonomal matrix with eigenvector (1,...,1)^T
     189PDAF     --- Floating point parameters (Array param_real) ---
     190PDAF       param_real(1): Forgetting factor (usually >0 and <=1), required
     191PDAF     --- Further parameters ---
     192PDAF       n_modeltasks: Number of parallel model integration tasks
     193PDAF           >=1 for subtypes 0 and 1; not larger than total number of processors
     194PDAF           =1 required for subtypes 2 and 3
     195PDAF       screen: Control verbosity of PDAF
     196PDAF           0: no outputs
     197PDAF           1: basic output (default)
     198PDAF           2: 1 plus timing output
     199PDAF           3: 2 plus debug output
     200PDAF     +++++++++ End of option overview for the ETKF ++++++++++
    228201}}}
    229202
    … …  
    232205
    233206{{{
    234         Available options:
    235 
    236         Sub-types (Parameter subtype)
    237           0: full ensemble integration;  apply T-matrix analogously to SEIK
    238           2: Fixed error space basis; analysis with T-matrix
    239           3: Fixed state covariance matrix; analysis with T-matrix
    240           5: Offline mode; analysis with T-matrix
    241 
    242         Integer parameters (Array param_int)
    243           param_int(1): Dimension of state vector (>0), required
    244           param_int(2): Ensemble size (>0), required
    245           param_int(3): Size of lag for smoothing
    246           param_int(4): not used
    247           param_int(5): Type of forgetting factor; optional, default: 0
    248               0: fixed forgetting factor
    249               1: adaptive forgetting factor for full domain (experimental)
    250               2: locally adaptive forgetting factor (experimental)
    251           param_int(6): Type of ensemble transformation matrix; optional, default: 0
    252               0: deterministic transformation
    253               2: use product of 0 with random orthonomal matrix with eigenvector (1,...,1)^T
    254 
    255         Floating point parameters (Array param_real)
    256           param_real(1): Forgetting factor (usually >0 and <=1), required
    257 
    258         Further parameters
    259           n_modeltasks: Number of parallel model integration tasks
    260               >=1 for subtypes 0 and 1; not larger than total number of processors
    261               =1 required for subtypes 2 and 3
    262           screen: Control verbosity of PDAF
    263               0: no outputs
    264               1: basic output (default)
    265               2: 1 plus timing output
    266               3: 2 plus debug output
    267 
    268         +++++++++ End of option overview for the LETKF ++++++++++
     207PDAF     Available options for LETKF:
     208PDAF     --- Sub-types (Parameter subtype) ---
     209PDAF       0: full ensemble integration;  apply T-matrix analogously to SEIK
     210PDAF       2: Fixed error space basis; analysis with T-matrix
     211PDAF       3: Fixed state covariance matrix; analysis with T-matrix
     212PDAF       5: Offline mode; analysis with T-matrix
     213PDAF     --- Integer parameters (Array param_int) ---
     214PDAF       param_int(1): Dimension of state vector (>0), required
     215PDAF       param_int(2): Ensemble size (>0), required
     216PDAF       param_int(3): Size of lag for smoothing
     217PDAF       param_int(4): not used
     218PDAF       param_int(5): Type of forgetting factor; optional, default: 0
     219PDAF           0: fixed forgetting factor
     220PDAF           1: adaptive forgetting factor for full domain (experimental)
     221PDAF           2: locally adaptive forgetting factor (experimental)
     222PDAF       param_int(6): Type of ensemble transformation matrix; optional, default: 0
     223PDAF           0: deterministic transformation
     224PDAF           2: use product of 0 with random orthonomal matrix with eigenvector (1,...,1)^T
     225PDAF     --- Floating point parameters (Array param_real) ---
     226PDAF       param_real(1): Forgetting factor (usually >0 and <=1), required
     227PDAF     --- Further parameters ---
     228PDAF       n_modeltasks: Number of parallel model integration tasks
     229PDAF           >=1 for subtypes 0 and 1; not larger than total number of processors
     230PDAF           =1 required for subtypes 2 and 3
     231PDAF       screen: Control verbosity of PDAF
     232PDAF           0: no outputs
     233PDAF           1: basic output (default)
     234PDAF           2: 1 plus timing output
     235PDAF           3: 2 plus debug output
     236PDAF     +++++++++ End of option overview for the LETKF ++++++++++
    269237}}}
    270238
    … …  
    273241
    274242{{{
    275         Available options:
    276 
    277         Sub-types (Parameter subtype)
    278           0: Standard implementation with ensemble integration
    279           2: Fixed error space basis
    280           3: Fixed state covariance matrix
    281           5: Offline mode
    282 
    283         Integer parameters (Array param_int)
    284           param_int(1): Dimension of state vector (>0), required
    285           param_int(2): Ensemble size (>0), required
    286           param_int(3): Size of lag for smoothing
    287           param_int(4): not used
    288           param_int(5): Type of forgetting factor; optional, default: 0
    289               0: fixed forgetting factor
    290               1: adaptive forgetting factor (experimental)
    291           param_int(6): Type of ensemble transformation matrix; optional, default: 0
    292               0: deterministic omega
    293               1: random orthonormal omega orthogonal to (1,...,1)^T
    294               2: use product of 0 with random orthonomal matrix with eigenvector (1,...,1)^T
    295                  (experimental; for random transformations, 0 or 1 are recommended)
    296           param_int(7): Type of transformation matrix square root; optional, default: 0
    297               0: symmetric square root
    298               1: Cholesky decomposition
    299 
    300         Floating point parameters (Array param_real)
    301           param_real(1): Forgetting factor (usually >0 and <=1), required
    302 
    303         Further parameters
    304           n_modeltasks: Number of parallel model integration tasks
    305               >=1 for subtypes 0 and 1; not larger than total number of processors
    306               =1 required for subtypes 2 and 3
    307           screen: Control verbosity of PDAF
    308               0: no outputs
    309               1: basic output (default)
    310               2: 1 plus timing output
    311               3: 2 plus debug output
    312 
    313         +++++++++ End of option overview for the ESTKF  ++++++++++
     243PDAF     Available options for ESTKF:
     244PDAF     --- Sub-types (Parameter subtype) ---
     245PDAF       0: Standard implementation with ensemble integration
     246PDAF       2: Fixed error space basis
     247PDAF       3: Fixed state covariance matrix
     248PDAF       5: Offline mode
     249PDAF     --- Integer parameters (Array param_int) ---
     250PDAF       param_int(1): Dimension of state vector (>0), required
     251PDAF       param_int(2): Ensemble size (>0), required
     252PDAF       param_int(3): Size of lag for smoothing
     253PDAF       param_int(4): not used
     254PDAF       param_int(5): Type of forgetting factor; optional, default: 0
     255PDAF           0: fixed forgetting factor
     256PDAF           1: adaptive forgetting factor (experimental)
     257PDAF       param_int(6): Type of ensemble transformation matrix; optional, default: 0
     258PDAF           0: deterministic omega
     259PDAF           1: random orthonormal omega orthogonal to (1,...,1)^T
     260PDAF           2: use product of 0 with random orthonomal matrix with eigenvector (1,...,1)^T
     261PDAF              (experimental; for random transformations, 0 or 1 are recommended)
     262PDAF       param_int(7): Type of transformation matrix square root; optional, default: 0
     263PDAF           0: symmetric square root
     264PDAF           1: Cholesky decomposition
     265PDAF     --- Floating point parameters (Array param_real) ---
     266PDAF       param_real(1): Forgetting factor (usually >0 and <=1), required
     267PDAF     --- Further parameters ---
     268PDAF       n_modeltasks: Number of parallel model integration tasks
     269PDAF           >=1 for subtypes 0 and 1; not larger than total number of processors
     270PDAF           =1 required for subtypes 2 and 3
     271PDAF       screen: Control verbosity of PDAF
     272PDAF           0: no outputs
     273PDAF           1: basic output (default)
     274PDAF           2: 1 plus timing output
     275PDAF           3: 2 plus debug output
     276PDAF     +++++++++ End of option overview for the ESTKF  ++++++++++
    314277}}}
    315278
    … …  
    318281
    319282{{{
    320         Available options:
    321 
    322         Sub-types (Parameter subtype)
    323           0: Standard implementation with ensemble integration
    324           2: Fixed error space basis
    325           3: Fixed state covariance matrix
    326           5: Offline mode
    327 
    328         Integer parameters (Array param_int)
    329           param_int(1): Dimension of state vector (>0), required
    330           param_int(2): Ensemble size (>0), required
    331           param_int(3): Size of lag for smoothing
    332           param_int(4): not used
    333           param_int(5): Type of forgetting factor; optional, default: 0
    334               0: fixed forgetting factor
    335               1: adaptive forgetting factor for full domain (experimental)
    336               2: locally adaptive forgetting factor (experimental)
    337           param_int(6): Type of ensemble transformation matrix; optional, default: 0
    338               0: deterministic omega
    339               1: random orthonormal omega orthogonal to (1,...,1)^T
    340               2: use product of 0 with random orthonomal matrix with eigenvector (1,...,1)^T
    341                  (experimental; for random transformations, 1 is recommended)
    342           param_int(7): Type of transformation matrix square root; optional, default: 0
    343               0: symmetric square root
    344               1: Cholesky decomposition
    345 
    346         Floating point parameters (Array param_real)
    347           param_real(1): Forgetting factor (usually >0 and <=1), required
    348 
    349         Further parameters
    350           n_modeltasks: Number of parallel model integration tasks
    351               >=1 for subtypes 0 and 1; not larger than total number of processors
    352               =1 required for subtypes 2 and 3
    353           screen: Control verbosity of PDAF
    354               0: no outputs
    355               1: basic output (default)
    356               2: 1 plus timing output
    357               3: 2 plus debug output
    358 
    359         +++++++++ End of option overview for the LESTKF ++++++++++
    360 }}}
     283PDAF     Available options for LESTKF:
     284PDAF     --- Sub-types (Parameter subtype) ---
     285PDAF       0: Standard implementation with ensemble integration
     286PDAF       2: Fixed error space basis
     287PDAF       3: Fixed state covariance matrix
     288PDAF       5: Offline mode
     289PDAF     --- Integer parameters (Array param_int) ---
     290PDAF       param_int(1): Dimension of state vector (>0), required
     291PDAF       param_int(2): Ensemble size (>0), required
     292PDAF       param_int(3): Size of lag for smoothing
     293PDAF       param_int(4): not used
     294PDAF       param_int(5): Type of forgetting factor; optional, default: 0
     295PDAF           0: fixed forgetting factor
     296PDAF           1: adaptive forgetting factor for full domain (experimental)
     297PDAF           2: locally adaptive forgetting factor (experimental)
     298PDAF       param_int(6): Type of ensemble transformation matrix; optional, default: 0
     299PDAF           0: deterministic omega
     300PDAF           1: random orthonormal omega orthogonal to (1,...,1)^T
     301PDAF           2: use product of 0 with random orthonomal matrix with eigenvector (1,...,1)^T
     302PDAF              (experimental; for random transformations, 1 is recommended)
     303PDAF       param_int(7): Type of transformation matrix square root; optional, default: 0
     304PDAF           0: symmetric square root
     305PDAF           1: Cholesky decomposition
     306PDAF     --- Floating point parameters (Array param_real) ---
     307PDAF       param_real(1): Forgetting factor (usually >0 and <=1), required
     308PDAF     --- Further parameters ---
     309PDAF       n_modeltasks: Number of parallel model integration tasks
     310PDAF           >=1 for subtypes 0 and 1; not larger than total number of processors
     311PDAF           =1 required for subtypes 2 and 3
     312PDAF       screen: Control verbosity of PDAF
     313PDAF           0: no outputs
     314PDAF           1: basic output (default)
     315PDAF           2: 1 plus timing output
     316PDAF           3: 2 plus debug output
     317PDAF     +++++++++ End of option overview for the LESTKF ++++++++++
     318}}}