fips.problems.flux.FluxProblem

class fips.problems.flux.FluxProblem(obs, prior, forward_operator, modeldata_mismatch, prior_error, constant=None, round_index=6)

Atmospheric flux inversion problem.

Subclass of InverseProblem specialized for estimating spatial and temporal surface fluxes from observed atmospheric concentrations using a forward transport model. Supports multi-block state composition (e.g., fluxes + bias corrections).

Attributes

averaging_kernel	Averaging kernel matrix (A).
background	Background concentration.
concentration_error	Model-data mismatch covariance matrix for observations.
concentrations	Observed concentrations.
desroziers	Desroziers et al. (2005) diagnosed observation error covariance.
enhancement	The background-subtracted observation vector.
estimator	Return the fitted estimator (raises if not solved).
jacobian	Forward operator (Jacobian) mapping state to observations.
kalman_gain	Kalman gain matrix (K).
n_obs	Return number of observations.
n_state	Return number of state variables.
obs_index	Return the observation space index.
plot	Plotting interface for flux inversion results.
posterior	Posterior state estimate.
posterior_concentrations	Modelled concentrations from posterior fluxes.
posterior_error	Posterior error covariance.
posterior_flux_error	Posterior flux error covariance matrix after inversion.
posterior_fluxes	Posterior flux estimates after inversion.
posterior_obs	Modeled observations (H @ posterior).
prior_concentrations	Modelled concentrations from prior fluxes.
prior_flux_error	Inventory error covariance matrix.
prior_fluxes	Prior flux inventory.
prior_obs	Modeled observations using the prior (H @ prior).
state_index	Return the state space index.
obs	Observation vector.
prior	Prior state vector.
forward_operator	Forward operator mapping state space to observation space.
prior_error	Covariance matrix representing prior error.
modeldata_mismatch	Covariance matrix representing model-data mismatch (observation error).
constant	Optional constant term added to the forward model (e.g., background or bias).

Methods

from_file(path)	Load object from a pickle file.
get_block(component, block[, crossblock])	Get block from a component (Vector or Matrix).
solve([estimator])	Solve the flux inversion problem using specified estimator.
to_file(path)	Save object to a pickle file.