Version history¶
Semantics
Version numbering aims to follow semantic versioning. Therefore:
- New patch versions (e.g. 1.1.0 to 1.1.1) make minor changes that do not alter fuctionality or calculated results.
- New minor versions (e.g. 1.0.1 to 1.1.0) add new functionality, but will not break your code. They will not alter the results of calculations with default settings (except for in the hopefully rare case of correcting a bug or typo).
- New major versions (e.g. 1.1.1 to 2.0.0) may break your code and require you to rewrite things. They may significantly alter the results of calculations with default settings.
We will always add aliases for existing functions if their API is updated, to avoid unforseen breaking changes wherever possible.
Warning
Will (not) break your code refers only to the functions covered in this documentation.
For the main CO2SYS functions as imported with
import PyCO2SYS as pyco2
results = pyco2.sys(*args, **kwargs) # new Pythonic syntax
results = pyco2.CO2SYS(*args, **kwargs) # old MATLAB syntax
the only things that may change, in at least a minor version release, are:
- Additional inputs may be added to the
kwargs
, but always with default values such that the results do not change if they are not explicitly set. - Additional calculated variables may be returned in the output
results
.
The structure of the underlying modules and their functions is not yet totally stable and, for now, may change in any version increment. Such changes will be described in the release notes below.
1.8¶
Adds atmospheric pressure input for pCO2-fCO2-xCO2 interconversions and (from v1.8.2) optional hydrostatic pressure effect on CO2 solubility and fugacity. Uncertainty analysis updated for more reproducible results. Rigorously validated and fully documented in peer-reviewed journal article (Humphreys et al., 2022).
1.8.3 (16 February 2024)¶
Changes in v1.8.3
New features
- Added
"dlnfCO2_dT"
and"dlnpCO2_dT"
results, the theoretical effect of temperature on the natural log of ƒCO2 and pCO2. - Added the PLR18 parameterisation of the carbonic acid constants for sea-ice brines.
Default options
- Reverted default
opt_k_carbonic
to10
(i.e., LDK00) for consistency with the best practice guide.
Bug fixes
- Updated
pyco2.equilibria.p1atm.kH2CO3_NBS_MCHP73
(used foropt_k_carbonic
options6
and7
) to update any salinity values less than 10–16 to be 10–16, because zero salinities give a NaN for K2, which causes problems for Autograd. This should not make any practical difference, because the parameterisation is only valid for salinities above 19. - Added
opt_pressured_kCO2
to results dict and incorporated it correctly into the uncertainty propagation functions.
Technical
- Updated from building with setup.py to pyproject.toml.
- PyCO2SYS can now be installed with conda/mamba (via conda-forge).
1.8.2 (19 January 2023)¶
Changes in v1.8.2
New features
- Added
opt_pressured_kCO2
to enable pressure corrections for the fugacity factor and CO2 solubility constant following W74. These have been added to CO2SYS-MATLAB by Jon Sharp at the same time with consistent results (differences less than 10−4 %). These pressure corrections are not enabled by default, for consistency with previous versions.
Bug fixes
- Fixed logicals in
solve.core()
that had meant no results were returned for parameter type combinations7, 10
,7, 11
,8, 10
and8, 11
. - Updated Autograd version for compatibility with Python 3.11.
1.8.1 (18 May 2022)¶
Changes in v1.8.1
Breaking changes
- For consistency with other settings,
buffers_mode
kwarg key changed toopt_buffers_mode
and its values are now integers rather than strings.
New features
- Adds new
par1_type
/par2_type
options10
and11
for saturation states with respect to calcite and aragonite. - Adds KSK18 parameterisation for estimating total borate from salinity.
Dependencies
- Switched to Autograd v1.4.
1.8.0 (27 October 2021)¶
Changes in v1.8.0
New features
- Adds
pressure_atmosphere
andpressure_atmosphere_out
arguments, rather than assuming 1 atm total barometric pressure.
Behind-the-scenes improvements
- Adds additional constraint to the initial pH estimate for more robust results for the alkalinity-CO2 fugacity parameter pair.
- Difference derivatives for uncertainties now have a fixed step size for each argument, instead of scaling depending on arguments, for more reproducible results.
1.7¶
Adds new syntax to return equilibrium constants and total salts without needing to solve the full carbonate system. Fully documented in manuscript in review (Humphreys et al., 2021, Geosci. Model Dev. Discuss.).
1.7.1 (10 August 2021)¶
Changes in v1.7.1
Bug fixes
- Improved handling of zero-valued inputs.
- Adjusted
CO2SYS_wrap
to work with latest pandas release.
1.7.0 (13 May 2021)¶
Changes in v1.7.0
New features
- Can now run
pyco2.sys
with no carbonate system parameter arguments provided, to just return all the equilibrium constants etc. under the specified conditions. - Can also run
pyco2.sys
with only one carbonate system parameter argument. This does not solve the carbonate system, but does calculate all that can be calculated with that parameter. - Added carbonic acid constants parameterisation of SB21.
- Added bisulfate dissociation constant parameterisation of WM13/WMW14.
- Added spreadsheet-to-spreadsheet function
pyco2.ezio
(with thanks to Daniel Sandborn). - Integrated uncertainty propagation into the main
pyco2.sys
function and expanded its capabilities.
Internal updates
- Switched default first-guess pH for solving from the alkalinity-carbonate ion parameter pair at low alkalinity from 10 to 3.
- Renamed various internal functions and variables for better consistency with the Pythonic
pyco2.sys
i/o syntax. - Removed the
PyCO2SYS.test
module, instead defining the round-robin test functions it contained directly in the test suite. - Added various internal settings for testing and validation against older CO2SYS-MATLAB versions.
- Adjust aqueous CO2 calculation for better consistency with CO2SYS-MATLAB (but negligible changes in the results).
- Can now use
PyCO2SYS.hello()
to find version number and credits (alias forPyCO2SYS.say_hello()
). - The final component of DIC (or DIC itself) to be calculated is now always computed by difference from the known components.
- Various functions in
convert
module renamed.
Validation
- Rigorous validation against various CO2SYS-MATLAB versions performed, as described in forthcoming PyCO2SYS manuscript (Humphreys et al., in prep.).
Bug fixes
par1
,par2
,par1_type
andpar2_type
arguments now always get broadcasted to the maximum size, even if they are scalar.- Erroneous
"k_phosphate_*"
keys corrected to"k_phosphoric_"
. - Override values for equilibrium constants under output conditions now assigned correctly.
- Fixed minor errors in initial pH estimates when solving from alkalinity and either DIC or [CO_2(aq)].
1.6¶
Adds extra alkalinity components with arbitrary pK values.
1.6.0 (26 October 2020)¶
Changes in v1.6.0
Bug fixes
- Updates the total alkalinity equation to fix minor error in pH scale conversions inherited from CO2SYS-MATLAB (see related note in v1.5.0 release notes).
New inputs and outputs
- Enables inputting total molalities and equilibrium constants for up to two additional contributors to total alkalinity.
- Full chemical speciation returned in the output dict of
pyco2.sys
, not just the alkalinity components as before.
New syntax
- Adds
sys
as an alias forCO2SYS_nd
at the top level. Recommended Python-style syntax is thus nowpyco2.sys
.
1.5¶
Introduces a more Pythonic top-level function that accepts multidimensional arguments and that only returns results at "output" conditions if explicitly specified.
1.5.1 (30 July 2020)¶
Changes in v1.5.1
Bug fixes
- Switched
dx
-scaling function inPyCO2SYS.uncertainties
to usenumpy.nanmedian
instead ofnumpy.median
. - Fixed
PyCO2SYS.uncertainties.propagate_nd
bug that prevented calculations on non-scalar arguments.
1.5.0 (29 July 2020)¶
Changes in v1.5.0
New top-level functions
- Adds
PyCO2SYS.CO2SYS_nd
top-level function with a more Pythonic interface and with NumPy broadcasting of n-dimensional inputs. - In
PyCO2SYS.CO2SYS_nd
, results at "output" conditions are only calculated if output temperature or pressure is provided. - Adds corresponding
PyCO2SYS.uncertainty.forward_nd
andPyCO2SYS.uncertainty.propagate_nd
functions for uncertainty propagation.
Alternative calculations
- New alkalinity equation fixing pH scale conversion bug inherited from CO2SYS-MATLAB is available, but not yet implemented by default.
Extra arguments and results
- Solubility constants for aragonite and calcite available directly as outputs from
PyCO2SYS.CO2SYS
andPyCO2SYS.CO2SYS_nd
- Explicit values for the solubility constants can be given as arguments to override the default internal calculation.
Internal reorganisation
- All 2-to-3 functions in
PyCO2SYS.solve.get
now have a more consistent set of inputs. - Revised pH scale conversion functions for consistency and added tests for their internal accuracy.
- Switched preallocations to use
np.shape
notnp.size
in preparation for working with n-dimensional inputs. - Updated style to import the whole NumPy module as
np
instead of individual functions separately. - Converted
PyCO2SYS.engine
to a sub-module.
1.4¶
Enables uncertainty propagation with forward finite-difference derivatives.
1.4.3 (16 July 2020)¶
Changes in v1.4.3
Bug fixes
- Corrected missing a pH scale conversion in SLH20 option for carbonic acid dissociation. Any calculations with this option in PyCO2SYS v1.4.1 or v1.4.2 should be updated!
Validation
- Results validated against new release candidate version of CO2SYS-MATLAB v3.
New API
- New wrapper function with the same input order and default gas constant option as the new CO2SYS-MATLAB v3 available in
PyCO2SYS.api.CO2SYS_MATLABv3
.
Internal reorganisation
_approx
function inputs inPyCO2SYS.solve.delta
updated to match the exact Autograd functions for easier switching.
1.4.2 (9 July 2020)¶
Changes in v1.4.2
Bug fixes
- Swapped order of
totals
andKs
arguments for all functions inPyCO2SYS.solve.delta
for consistency with other modules. - Inverted the alkalinity-pH residual equations in
PyCO2SYS.solve.delta
.
Reorganisation
- Broke out some parts of
PyCO2SYS.equilibria.assemble
into separate functions.
1.4.1 (1 July 2020)¶
Changes in v1.4.1
Extra calculation options
- Added the 2018 CODATA value for the universal gas constant R as an option for consistency with forthcoming CO2SYS-MATLAB v3. The original DOEv2 version remains default.
- Added the SLH20 equations as option
16
for the carbonic acid dissociation constants.
1.4.0 (9 June 2020)¶
Changes in v1.4.0
New features
- Added
uncertainty
module with functions to evaluate derivatives of PyCO2SYS outputs with respect to inputs, along with corresponding documentation. - Specific input values can optionally be provided for all total concentrations and equilibrium constants that are estimated internally from salinity, temperature and pressure.
General improvements
- Added basic sanity checking to prevent some invalid marine carbonate system parameter input values.
- Nutrient concentrations have always been set to zero internally for
K1K2CONSTANTS
options6
and8
, and salinity too for8
, regardless of the input values. This is now reflected in the output values of these variables in theCO2dict
.
New outputs
- Substrate:inhibitor ratio (SIR) of B15, calculated with
SIratio
in new modulebio
. - Inputs
PAR1
andPAR2
. - The "Peng correction" factor.
- The fugacity factor for converting between CO2 partial pressure and fugacity.
- The activity coefficient of the H+ ion for NBS pH scale conversions.
Validation
- Calculations compare very favourably against the forthcoming CO2SYS for MATLAB v3 - see Validation for discussion of the results.
1.3¶
Adds bicarbonate ion and aqueous CO2 as inputs from which the carbonate system can be solved. Continues to reorganise code behind the scenes. Makes almost everything Autograd-able and uses this approach to calculate buffer constants. Validates results against CO2SYS for MATLAB.
1.3.0 (1 May 2020)¶
Changes in v1.3.0
New features
- Added bicarbonate ion (type
7
) and aqueous CO2 (type8
) as options for known input marine carbonate system variables. - Added module
test
with functions to perform internal consistency checks onPyCO2SYS.CO2SYS
calculations and compare results with those from other sources. - Added module
api
with a wrapper forPyCO2SYS.CO2SYS
to allow inputs as Pandas Series and/or Xarray DataArrays.
Improved calculations
- The Revelle factor and all other buffer factors added in v1.2 are now evaluated using automatic differentiation, which means that the effects of all equilibrating species are taken into account.
- The original, non-automatic functions that do not account for nutrient effects are still available in
buffers.explicit
. - Can switch between calculation methods using new optional input
opt_buffers_mode
.
- The original, non-automatic functions that do not account for nutrient effects are still available in
- Corrected Revelle factor calculations:
- Added missing "Peng correction" to Revelle factor calculation at output conditions. Note that this correction is currently also missing from CO2SYS for MATLAB!
- Decreased DIC perturbation size for more accurate finite-difference "explicit" evaluation.
- Finite-difference calculation now references the correct DIC value.
- Implemented better initial guesses for pH in all iterative solvers in
solve.get
following M13 and OE15. - Switched to using exact slopes in iterative solvers in
solve.get
, evaluated using Autograd in new submodulesolve.delta
. - Updated entire package to be Autograd-able.
- Return NaN instead of negative DIC if an impossible pH-alkalinity combination is given as input (i.e. pH is too high).
- Return NaN where DIC and one of its components is given if the component is impossibly large.
Internal reorganisation
- Major internal reorganisation that is probably not fully captured in these notes.
- Renamed modules:
assemble
is nowengine
.concentrations
is nowsalts
.extra
is nowbuffers.explicit
.
- Module
equilibria
now contains sub-modules:p1atm
for calculating constants at atmospheric pressure.pcx
for determining pressure correction factors.pressured
for calculating constants at given pressure.
- Module
solve
now contains sub-modules:initialise
to generate first-guess estimates of pH for the TA-pH solvers.get
to calculate a new system variable from various input pairs.
- Added module
solubility
for mineral solubility calculations. - Relocated
_CaSolubility
function from root tosolubility.CaCO3
.- Separated out its internal calculations into a set of subfunctions also in the
solubility
module. - Added calcium molinity
TCa
as an input, instead of being evaluated internally.
- Separated out its internal calculations into a set of subfunctions also in the
- Added calcium molininty
TCa
(estimated from salinity) into the mainCO2dict
output fromPyCO2SYS.CO2SYS
. - Relocated
_RevelleFactor
function from root tobuffers.RevelleFactor
. - Relocated
_FindpHOnAllScales
function from root toconvert.pH2allscales
. - Added module
constants
for storing values of universal physical constants. - Lists of equilibrium constants and total concentrations now passed around internally as dicts, for safety.
- Total sulfate and bisulfate dissociation constant renamed from
TS
andKS
toTSO4
andKSO4
internally to avoid confusion with sulfide species. - The as-close-as-possible MATLAB clone in
PyCO2SYS.original
no longer produces a dict but just the originalDATA
,HEADERS
andNICEHEADERS
outputs.
Miscellaneous
- Documentation substantially expanded and switched to using Material for MkDocs.
- Validation section added with internal consistency checks and an external comparison of PyCO2SYS calculations with CO2SYS for MATLAB.
- All code now formatted with Black (except for module
original
). - Version number now reported with
PyCO2SYS.say_hello()
in homage to the original MS-DOS program.
1.2¶
Calculates a wider variety of chemical buffer factors.
1.2.1 (9 April 2020)¶
Changes in v1.2.1
1.2.0 (8 April 2020)¶
Changes in v1.2.0
- Added module
extra
containing functions to calculate variables not included in CO2SYS for MATLAB: buffers_ESM10
calculates the buffer factors of ESM10, corrected for the typos noted by RAH18.bgc_isocap
calculates the "exact" isocapnic quotient of HDW18, Eq. 8.bgc_isocap_approx
calculates the approximate isocapnic quotient of HDW18, Eq. 7.psi
calculates the \psi factor of FCG94.- Added all functions in
extra
to theCO2dict
output of the mainCO2SYS
function, and documented in the Github repo README.
1.1¶
Adds optional inputs of total ammonium, hydrogen sulfide, and carbonate ion molinities for consistency with forthcoming MATLAB "v1.21".
1.1.1 (20 March 2020)¶
Changes in v1.1.1
- Removed unnecessary
WhoseTB
input toassemble.equilibria
.
1.1.0 (19 March 2020)¶
Changes in v1.1.0
- Updated pH-solving iterative functions so that iteration stops separately for each row once it reaches the tolerance threshold.
- Extracted all functions for solving the CO2 system into a separate module (
solve
). - Extracted other key subfunctions into module
assemble
. - Added total ammonium (
NH3
) and hydrogen sulfide (H2S
) concentrations as optional inputs to be included in the alkalinity model. - Added optional input to choose between different equations for hydrogen fluoride dissociation constant (
KFCONSTANT
). - Added functions to enable carbonate ion as an input carbonate system variable.
- Output is now only the
CO2dict
dict, not the originalDATA
,HEADERS
andNICEHEADERS
. - Eliminated all global variables throughout the entire program.
1.0¶
1.0.1 (28 February 2020)¶
Starts to make things more Pythonic.
Changes in v1.0.1
- Extracted all equations for concentrations and equilibrium constants into functions in separate modules (
concentrations
andequilibria
). - Eliminated all global variables from the
_Constants
function. - Moved the as-close-as-possible version into module
original
. The defaultfrom PyCO2SYS import CO2SYS
now imports the more Pythonic implementation.
1.0.0 (3 February 2020)¶
An as-close-as-possible clone of MATLAB CO2SYS v2.0.5.
Release notes for v1.0.0
- The first output
DICT
is new: a dict containing a separate entry for each variable in the original outputDATA
, with the keys named following the original outputHEADERS
. - The output
DATA
is transposed relative to the MATLAB version because Numpy is row-major while MATLAB is column-major. - Every combination of input options was tested against the MATLAB version with no significant differences (i.e. all differences can be attributed to floating point errors).