Development of stochastically perturbed parameterization scheme for the Noah Land Surface Model with the optimized random forcing parameters using the micro-genetic algorithm

<p>The ensemble data assimilation system is beneficial to represent the initial uncertainties and flow-dependent background error covariance (BEC). In particular, the inevitable model uncertainties can be expressed by ensemble spread, that is the standard deviation of ensemble BEC. However, the ensemble spread generally suffers from under-estimated problems. To alleviate this problem, recent studies employed stochastic perturbation schemes to increases the ensemble spreads by adding the random forcing in the model tendencies (i.e., physical or dynamical tendencies) or parameterization schemes (i.e., PBL, convective scheme, etc.). In this study, we focus on the near-surface uncertainties which are affected by the interactions between the land and atmosphere process. The land surface model (LSM) provides various fluxes as the lower boundary condition to the atmosphere, influencing the accuracy of hourly-to-seasonal scale weather forecasting, but the surface uncertainties were not much addressed yet. In this study, we developed the stochastically perturbed parameterization (SPP) scheme for the Noah LSM. The Weather Research and Forecasting (WRF) ensemble system is used for regional weather forecasting over East Asia, especially over the Korean Peninsula. As a testbed experiment with the newly-developed Noah LSM-SPP system, we first perturbed the soil temperature — a crucial variable for the near-surface forecasts by affecting sensible heat fluxes, land surface skin temperature and surface air temperature, and hence lower-tropospheric temperature. Here, the random forcing used in perturbation is made by the tuning parameters for amplitude, length scale, and time scales: they are commonly determined empirically by trial and error. In order to find optimal tuning parameter values, we applied a global optimization algorithm — the micro-genetic algorithm (micro-GA) — to achieve the smallest root-mean-squared errors. Our results indicate that optimization of the random forcing parameters contributes to an increase in the ensemble spread and a decrease in the ensemble mean errors in the near-surface and lower-troposphere uncertainties. Further experiments will be conducted by including soil moisture in the testbed.</p>

Representation of model error through process-based perturbations for ensemble prediction : application to turbulence and shallow convection parametrisations

<p>The boundary layer is the place of many complex physical processes spanning various time and space scales, part of which need to be parametrised in NWP models. These parametrisations are known sources of uncertainty in the models, due to the difficulty of accurately representing the processes, and the resulting simplifications and approximations that have to be done. Model uncertainty is part of what ensemble prediction systems seek to represent. This can be achieved in particular by using stochastic perturbation methods, where noise is introduced during model computations to change its state and produce different simulations. Well-known and widely used perturbation schemes like the Stochastically Perturbed Parametrisation Tendencies (SPPT) scheme have shown their effectiveness and their interest in building ensembles. However, part of the model uncertainty is not yet well represented in current ensemble systems, while some of the assumptions made by SPPT can be questioned. This argues for a diversity of approaches to represent model errors. In this active research field, alternative perturbation methods are investigated, such as the Stochastically Perturbed Parametrisations (SPP) method, or other methods focusing on the perturbation of particular physical processes. The work presented here focuses on the last ones. Based on two examples of methods published in the literature, perturbations have been applied to the turbulence and shallow convection parametrisation schemes of the mesoscale NWP model Arome from Météo-France. The perturbation of turbulence is based on the use of subgrid-scale variances to regulate the amplitude of an additive noise, while shallow convection is perturbed through a stochastic closure condition of the scheme. A simplified 1D framework has been used, in order to assess the ability of the method to produce an ensemble with sufficient dispersion and to compare its results with those from existing methods like SPPT.</p>

Sensitivity of modeled microphysics to stochastically perturbed parameters

<p>This study investigates sensitivity of  cloud and precipitation parameterized microphysics  to stochastic representation of parameter uncertainty as formulated by the stochastically perturbed parameterization (SPP) scheme.  SPP is applied to multiple microphysical parameters within a lagrangian column model, used in several prior published studies to characterize  parameter uncertainty by means of multivariate nonlinear inversions using remote sensing observations. The 1D column microphysics model is forced with prescribed time-varying profiles of temperature, humidity and vertical velocity.  This modeling framework allows for investigation of the effect of changes in model physics parameters on the model output in isolation from any feedback to the cloud-scale dynamics.</p><p>The test case selected in this study of an idealized representation of mid-latitude squall-line convection is the same as in the prior studies. This enabled using the estimates of multi-parameter distributions from the inversions in the prior studies as the basis for setting the second-moment statistics in the SPP scheme implementation. Additionally impacts of the non-stochastic and stochastic multi-parameter representation of parameterization uncertainty on the microphysics model solution could be directly compared.</p><p>The sensitivity experiments with the SPP scheme involve ensemble simulations where each member is evolved with a different stochastic sequence of parameter perturbations, as is done in the standard practice with this scheme.  The experiments explore impacts of using different decorrelation times and different estimates of second moment statistics for the parameter perturbations.  These include uncorrelated perturbations between the parameters for several values of variance for each parameter and correlated perturbations based on multi-parameter empirical statistical distributions from the prior studies.  The selection of physical parameters for the perturbations is based on the significance of their impacts derived from the prior studies . </p><p>The results are evaluated in terms of changes to the ensemble mean and variance of time evolving profiles of hydrometeor mass quantities, the microphysics processes within the model as well as in terms of the simulated column integral microphysics-sensitive satellite-based  observables. The latter include PR (Precipitation Rate) , LWP (Liquid Water Path), IWP (Ice water path), TOA-LW and TOA-SW (-Long and -Short Wave, respectively).  In each experiment six parameters were perturbed.</p><p>The analyses performed so far indicate a high sensitivity of the microphysics model to the SPP scheme. The ensemble simulations with the standard uncorrelated parameter perturbations exhibit a significant bias relative to the control simulation which uses the unperturbed parameters.  For the selected test case the skewness toward small parameter values in the SPP sampling based on the underlying log-normal distributions leads to less precipitating ice and more precipitating liquid and accumulated precipitation. The response is due to nonlinear relationships between the parameters and modeled microphysics output. The changes in microphysics output result in large mean changes in PR, LWP, IWP, TOA- LW and SW, suggesting a potential for using these and other microphysics sensitive satellite observations to evaluate and if needed correct properties of the underlying sampling distribution in the stochastic scheme.  Further analyses will be presented at the conference.</p>

Progress in ensemble forecasting and verification methodologies at ECMWF

<p>This talk focusses on progress in ensemble forecasting methodology (Part I) and ensemble verification methodology (Part II).</p><p>Operational ECMWF ensemble forecasts are global predictions from days to months ahead. At all forecast ranges, model uncertainties are represented stochastically with the Stochastically Perturbed Parametrization Tendency scheme (SPPT). Recently, considerable progress has been made in developing the Stochastically Perturbed Parametrization scheme (SPP). The SPP scheme offers improved physical consistency by naturally preserving the local conservation properties for energy and moisture of the unperturbed version of the corresponding parametrization. In contrast, the SPPT scheme lacks such local conservation properties, mainly because the scheme does not perturb fluxes at the surface and at the top of the atmosphere consistently with the tendency perturbations in the column.</p><p>NWP research and development relies on scoring rules to judge whether or not a change to the forecast systems results in better ensemble forecasts. A new tool will be presented that can improve the understanding of score differences between sets of forecasts for a widely used proper score, the Continuous Ranked Probability Score (CRPS). An analytical expression has been derived for the CRPS when a homogeneous Gaussian (hoG) forecast-observation distribution is considered. This leads to an approximation of the CRPS when actual verification data are considered, which deviate from a homogeneous Gaussian distribution. The hoG approximation of the CRPS permits a useful decomposition of score differences. The methodology will be illustrated with verification data for medium-range weather forecasts.</p>