Model Forecast Errors Research Articles

Instrument Choice, Carbon Emissions, and Information

This article examines the consequences of a previously unnoticed difference between pollutant cap-and-trade and pollution taxes. Implementation of cap-and-trade relies on a forecast of future emissions, implementation of a pollution tax does not. Realistic policy designs using either regulatory instrument almost always involve a phase-in over time to avoid economic disruption. Cap-and-trade accomplishes this phase-in via a limit on emissions that falls gradually below the forecast of future pollutant emissions. Emissions taxation accomplishes the same via a gradually increasing levy on pollution. Because of the administrative complexity of establishing an emissions trading market, cap-and-trade programs typically require between 3 and 5 years lead time before imposing obligations on emitters. I present new evidence showing that forecast error over this timeframe for United States energy related carbon dioxide emissions from the Department of Energy’s energy model – the model used for policy design by Congress and the EPA – is biased and imprecise to such a degree as to make such use impractical. Forecast emissions are insufficiently accurate to allow for creation of a reliable or predictable market signal to incentivize emission reductions. By contrast, carbon taxes, because they do not depend upon a baseline emissions forecast, create a relatively clear level of policy stringency. This difference matters because policies that end up weaker than intended face low odds for strengthening while those that end up stronger than intended are likely to be weakened. The political asymmetry combined with actual model forecast errors leads to bias in favor of suboptimal, weak, policies for cap and trade. This is a serious concern if, as is usually the case, a cap is set based on political bargaining rather than on an optimal balancing of abatement costs and avoided climate damage. By contrast, the same model bias would lead to more environmentally effective than forecast carbon taxes but without the political consequences created by price volatility, were such programs to be implemented in the US. Thus while theory tells us that cap-and-trade and carbon taxes can be equivalent, imperfect information leads to suboptimal environmental performance of emissions trading, relative to carbon taxation policies. Policy makers should weigh these practical, information related concerns when considering approaches to controlling emissions of greenhouse gases.

Assimilation of high frequency radar data into a shelf sea circulation model

High Frequency (HF) radar current data is assimilated into a shelf sea circulation model based on optimal interpolation (OI) method. The purpose of this work is to develop a real-time computationally highly efficient assimilation method to improve the forecast of shelf current. Since the true state of the ocean is not known, the specification of background error covariance is arduous. Usually, it is assumed or calculated from an ensemble of model states and is kept in constant. In our method, the spatial covariances of model forecast errors are derived from differences between the adjacent model forecast fields, which serve as the forecast tendencies. The assumption behind this is that forecast errors can resemble forecast tendencies, since variances are large when fields change quickly and small when fields change slowly. The implementation of HF radar data assimilation is found to yield good information for analyses. After assimilation, the root-mean-square error of model decreases significantly. Besides, three assimilation runs with variational observation density are implemented. The comparison of them indicates that the pattern described by observations is much more important than the amount of observations. It is more useful to expand the scope of observations than to increase the spatial interval. From our tests, the spatial interval of observation can be 5 times bigger than that of model grid.

Retrieval of tropical cyclone statistics with a high‐resolution coupled model and data

AbstractWhen observations are assimilated into a high‐resolution coupled model, a traditional scheme that preferably projects observations to correct large‐scale background tends to filter out small‐scale cyclones. Here we separately process the large‐scale background and the small‐scale perturbations with low‐resolution observations for reconstructing historical cyclone statistics in a cyclone‐permitting model. We show that by maintaining the interactions between small‐scale perturbations and successively corrected large‐scale background, a model can successfully retrieve the observed cyclone statistics that in return improve estimated ocean states. The improved ocean initial conditions together with the continuous interactions of cyclones and background flows are expected to reduce model forecast errors. Combined with convection‐permitting cyclone initialization, the new high‐resolution model initialization along with the progressively advanced coupled models should contribute significantly to the ongoing research on seamless weather‐climate predictions.

Characteristics of forecast errors in the National Climate Center atmospheric general circulation model in winter

By using the National Climate Center atmospheric general circulation model (BCC_AGCM) for the second generation monthly dynamic extended range system and the NCEP/NCAR reanalysis data accumulated during 1982-2010, the model forecast errors in winter temperature prediction and their relationship with external forcing factors are analyzed. Result indicates that the model can well reproduce the variation trends of winter temperature over Eurasia region, and the main interannual variability pattern of winter temperature over East Asian monsoon domain can be successfully presented. It reveals that the model has acceptable performances in winter temperature prediction over East Asian monsoon domain, while the forecast errors are still considerable. In spatial distribution, the forecast errors over lands are larger than those over oceans. Meanwhile, the errors at high latitudes are larger than at low latitudes, and they are also closely related to altitude. The principal components of forecast errors have significant association with the sea surface temperature and the sea ice concentration over some key regions, it is shown that the response capability to external forcings is insufficient. This may provide a valuable reference for improving the prediction skill of winter temperature for East Asian monsoon domain, in combination with the model response to external forcings.

Multicycle ensemble forecasting of sea surface temperature

AbstractA novel extension to time‐lagged ensemble forecasting called multicycle ensemble forecasting improves the independent sampling of forecast model errors. Multicycle is defined such that each forecast cycle is independent of the previous forecast cycle. For an M cycle system the background field for each cycle is from a model hindcast M cycles earlier. The model errors have a factor M longer period to grow compared with a sequential system; however, the increased independence in the forecast model errors provide weighted ensemble averages with greater skill and reliability over the 0 lag forecast and a good spread‐error relationship. This cost‐efficient technique is relevant to global ocean forecasting where an ensemble method is computationally prohibitive.

Statistical distribution for wind power forecast error and its application to determine optimal size of energy storage system

Accurate wind power forecast is an important tool for wind farm to participate in day-ahead or hours-ahead energy markets. However, forecast errors with any methodology are so large that they cannot be neglected. The forecast error needs to be analyzed individually for single wind farm to estimate the impact of this error on trading wind energy in electricity market. Although forecast error is always assumed as normal distribution, it can be demonstrated that it is not proper with a simple statistical analysis. In this paper, a mixed distribution is proposed based on laplace and normal distribution to model forecast errors associated with persistence forecast for single wind farm over multiple timescales. Then the proposed distribution is used to estimate the penalties for prediction errors in the electricity market. Energy storage system (ESS) can smooth the wind power output and make wind power more “dispatchable”. A probabilistic method is proposed to determine optimal size of ESS for wind farm in electricity markets. The results indicate that the proposed distribution and probabilistic method is efficient to find optimal size of ESS.

Analysis of the surface temperature and wind forecast errors of the NCAR-AirDat operational CONUS 4-km WRF forecasting system

Investigating the characteristics of model-forecast errors using various statistical and object-oriented methods is necessary for providing useful guidance to end-users and model developers as well. To this end, the random and systematic errors (i.e., biases) of the 2-m temperature and 10-m wind predictions of the NCAR-AirDat weather research and forecasting (WRF)-based real-time four-dimensional data assimilation (RTFDDA) and forecasting system are analyzed. This system has been running operationally over a contiguous United States (CONUS) domain at a 4-km grid spacing with four forecast cycles daily from June 2009 to September 2010. In the result an exceptionally useful forecast dataset was generated and used for studying the error properties of the model forecasts, in terms of both a longer time period and a broader coverage of geographic regions than previously studied. Spatiotemporal characteristics of the errors are investigated based on the 24-h forecasts between June 2009 and April 2010, and the 72-h forecasts between May and September 2010. It was found that the biases of both wind and temperature forecasts vary greatly seasonally and diurnally, with dependency on the forecast length, station elevation, geographical location, and meteorological conditions. The temperature showed systematic cold biases during the daytime at all station elevations and warm biases during the nighttime above 1,000 m above sea level (ASL), while below 600 m ASL cold biases occurred during the nighttime. The forecasts of surface wind speed exhibited strong positive biases during the nighttime, while the negative biases were observed in the spring and summer afternoons. The surface wind speed was mostly over-predicted except for the stations located between 1,000 and 2,100 m ASL, for which negative biases were identified for most forecast cycles. The highest wind-speed errors were found over the high terrain and near sea-level stations. The wind-direction errors were relatively large at the high-terrain elevation in the Rocky and Appalachian mountain ranges and the western coastal areas and the error structure exhibited notable diurnal variability.

Dryline Position Errors in Experimental Convection-Allowing NSSL-WRF Model Forecasts and the Operational NAM

Abstract This study evaluates 24-h forecasts of dryline position from an experimental 4-km grid-spacing version of the Weather Research and Forecasting Model (WRF) run daily at the National Severe Storms Laboratory (NSSL), as well as the 12-km grid-spacing North America Mesoscale Model (NAM) run operationally by the Environmental Modeling Center of NCEP. For both models, 0000 UTC initializations are examined, and for verification 0000 UTC Rapid Update Cycle (RUC) analyses are used. For the period 1 April–30 June 2007–11, 116 cases containing drylines in all three datasets were identified using a manual procedure that considered specific humidity gradient magnitude, temperature, and 10-m wind. For the 24-h NAM forecasts, no systematic east–west dryline placement errors were found, and the majority of the east–west errors fell within the range ±0.5° longitude. The lack of a systematic bias was generally present across all subgroups of cases categorized according to month, weather pattern, and year. In contrast, a systematic eastward bias was found in 24-h NSSL-WRF forecasts, which was consistent across all subgroups of cases. The eastward biases seemed to be largest for the subgroups that favored “active” drylines (i.e., those associated with a progressive synoptic-scale weather system) as opposed to “quiescent” drylines that tend to be present with weaker tropospheric flow and have eastward movement dominated by vertical mixing processes in the boundary layer.

Improving flow forecasting by error correction modelling in altered catchment conditions

AbstractDespite human is an increasingly significant component of the hydrologic cycle in many river basins, most hydrologic models are still developed to accurately reproduce the natural processes and ignore the effect of human activities on the watershed response. This results in non‐stationary model forecast errors and poor predicting performance every time these models are used in non‐pristine watersheds. In the last decade, the representation of human activities in hydrological models has been extensively studied. However, mathematical models integrating the human and the natural dimension are not very common in hydrological applications and nearly unknown in the day‐to‐day practice. In this paper, we propose a new simple data‐driven flow forecast correction method that can be used to simultaneously tackle forecast errors from structural, parameter and input uncertainty, and errors that arise from neglecting human‐induced alterations in conceptual rainfall–runoff models. The correction system is composed of two layers: (i) a classification system that, based on the current flow condition, detects whether the source of error is natural or human induced and (ii) a set of error correction models that are alternatively activated, each tailored to the specific source of errors. As a case study, we consider the highly anthropized Aniene river basin in Italy, where a flow forecasting system is being established to support the operation of a hydropower dam. Results show that, even by using very basic methods, namely if‐then classification rules and linear correction models, the proposed methodology considerably improves the forecasting capability of the original hydrological model. Copyright © 2013 John Wiley & Sons, Ltd.

Current Capability of Operational Numerical Models in Predicting Tropical Cyclone Intensity in the Western North Pacific

Abstract Forecasts of tropical cyclone (TC) intensity from six operational models (three global models and three regional models) during 2010 and 2011 are assessed to study the current capability of model guidance in the western North Pacific. The evaluation is performed on both Vmax and Pmin from several aspects, including the relative error, skill assessment, category score, the hitting rate of trend, and so on. It is encouraging to see that the models have had some skills in the prediction of TC intensity, including that two of them are better than a statistical baseline in Vmax at several lead times and three of them show some skill in intensity change. With dissipated cases included, all the models have skills in category and trend forecasting at lead times longer than 24 h or so. The model forecast errors are found to be significantly correlated with initial error and the observed initial intensity. A statistical calibration scheme for model forecasting is proposed based on such an attribute, which is more effective for Pmin than Vmax. The statistically calibrated model forecasts are important in setting up a skillful multimodel consensus, for either the mean or the statistically weighted mean. The Vmax forecasts converted from the calibrated Pmin consensus based on a statistical wind–pressure relationship show significant skill over the baseline and a skillful scheme is also proposed to deal with the delay of the model forecasts in operation.

High-resolution quantitative precipitation forecasts and simulations by the Cloud-Resolving Storm Simulator (CReSS) for Typhoon Morakot (2009)

Summary Typhoon Morakot (2009) struck Taiwan during 7–9 August 2009, and brought extreme rainfall up to 2855 mm and the worst damages in the past 50 years. The operational models showed deficiency and serious under-prediction in rainfall amount at real time. This study demonstrates that the Cloud-Resolving Storm Simulator (CReSS), a state-of-the-art, high-resolution model, at a grid size of 3 km and starting as early as 0000 UTC 4 August, can successfully simulate and reproduce the event with high accuracy, including the distribution and timing of heavy rainfall in Taiwan. In the simulation starting at 0000 UTC 6 August, for example, the threat scores for 24-h rainfall for 8 August (with extreme amounts >1450 mm) reach 0.8–0.4 even at thresholds of 100–500 mm. This result is only possible due to small track error and the phase-locking mechanism of the Taiwan topography to heavy rainfall. Furthermore, real-time forecast and hindcast integrations of the CReSS model show that high-quality quantitative precipitation forecasts (QPFs) with peak total amount 67–80% of the true value are also obtained from initial conditions at 0000 UTC 6 August, which is about 2 days prior to the beginning of the heaviest rainfall in southern Taiwan. In these integrations, typhoon track errors in the global model forecasts used as boundary conditions are the major error source that prevent more ideal QPF results before and at 1200 UTC 5 August. When properly configured, it is believed that other similar cloud-resolving models can achieve comparable performance. Thus, the importance of and potential benefits from deterministic high-resolution forecasts is stressed, which may give an extended lead-time when the track error is small. With potentially longer time window for emergency action just prior to extreme rainfall events when it matters the most, such forecasts may ultimately lead to reduced losses in lives and properties.

Diagnosing Forecast Errors in Tropical Cyclone Motion

Abstract This paper reports on the development of a diagnostic approach that can be used to examine the sources of numerical model forecast error that contribute to degraded tropical cyclone (TC) motion forecasts. Tropical cyclone motion forecasts depend upon skillful prediction of the environment wind field, and by extension, the synoptic-scale weather systems nearby the TC. While previous research suggests that the deep-layer mean (DLM) steering flow typically approximates the actual TC motion, it is shown that the motion of even mature TCs can depart from the DLM steering flow. An optimal environmental steering flow is defined, which varies the vertical extent of the steering layer and the radius over which TC vorticity and divergence are removed. Errors in predicted TC motion are quantified using a diagnostic equation that accounts for not only differences in the synoptic-scale flow, but also differences in the depth and radius used to define the steering flow. Differences in the latter two parameters are interpreted in terms of errors in predicted TC structure or errors in proximate mesoscale flow features. Results from an analysis of 24-h forecasts from the Advanced Hurricane Weather Research and Forecasting Model during the 2008–10 North Atlantic TC seasons show that forecast motion errors are dominated by errors in the environment wind field. Contributions from other terms are occasionally large and are interpreted from a vorticity perspective. The utility of this new diagnostic equation is that it can be used to assess TC motion forecasts from any numerical modeling system.

Determining Key Model Parameters of Rapidly Intensifying Hurricane Guillermo (1997) Using the Ensemble Kalman Filter

Abstract In this work the authors determine key model parameters for rapidly intensifying Hurricane Guillermo (1997) using the ensemble Kalman filter (EnKF). The approach is to utilize the EnKF as a tool only to estimate the parameter values of the model for a particular dataset. The assimilation is performed using dual-Doppler radar observations obtained during the period of rapid intensification of Hurricane Guillermo. A unique aspect of Guillermo was that during the period of radar observations strong convective bursts, attributable to wind shear, formed primarily within the eastern semicircle of the eyewall. To reproduce this observed structure within a hurricane model, background wind shear of some magnitude must be specified and turbulence and surface parameters appropriately specified so that the impact of the shear on the simulated hurricane vortex can be realized. To identify the complex nonlinear interactions induced by changes in these parameters, an ensemble of model simulations have been conducted in which individual members were formulated by sampling the parameters within a certain range via a Latin hypercube approach. The ensemble and the data, derived latent heat and horizontal winds from the dual-Doppler radar observations, are utilized in the EnKF to obtain varying estimates of the model parameters. The parameters are estimated at each time instance, and a final parameter value is obtained by computing the average over time. Individual simulations were conducted using the estimates, with the simulation using latent heat parameter estimates producing the lowest overall model forecast error.

First outcomes from the CNR-ISAC monthly forecasting system

Abstract. A monthly probabilistic forecasting system is experimentally operated at the ISAC institute of the National Council of Research of Italy. The forecasting system is based on GLOBO, an atmospheric general circulation model developed at the same institute. The model is presently run on a monthly basis to produce an ensemble of 32 forecasts initialized with GFS-NCEP perturbed analyses. Reforecasts, initialized with ECMWF ERA-Interim reanalyses of the 1989–2009 period, are also produced to determine modelled climatology of the month to forecast. The modelled monthly climatology is then used to calibrate the ensemble forecast of daily precipitation, geopotential height and temperature on standard pressure levels. In this work, we present the forecasting system and a preliminary evaluation of the model systematic and forecast errors in terms of non-probabilistic scores of the 500-hPa geopotential height. Results show that the proposed forecasting system outperforms the climatology in the first two weeks of integrations. The adopted calibration based on weighted bias correction is found to reduce the systematic and the forecast errors.

Comment

Comment

Application of Non-Linear Partial Least-Squares Regression in the Prediction of Carbonization Depth of Concrete

Based on partial least-squares multinomial regression, this paper had a prediction on carbonization depth of conctete. Taking water-cement ratio (i.e. water-ash ratio), cement content (i.e. application amount of cement material), and exposure time of concrete as independent variables , and carbonization depth of conctete as dependent variable , the forecast model of carbonization depth of conctete was obtained. It was found that, Press residual value decreased with the increase of number of latent variables, and the number of latent variables were three by Press residual value versus number of latent variables. The normal regression coefficient of exposure time of concrete was the largest wiithin three influence factors, this indicated that the influence of exposure time of concrete was largest to conctete carbonization depth; The determination coefficient of forecast model obtained in this paper was 0.9940, the error of forecast model was . The following conclusion can be drawn that, the model is accurate and credible, and the partial least-squares multinomial regression is a eximious non-linear method, and it is worthy to spread its application in the forecast analysis of conctete carbonization depth.

How bias-correction can improve air quality forecasts over Portugal

Abstract Currently three air quality modelling systems operate routinely with high-resolution over mainland Portugal for forecasting purposes, namely MM5-CHIMERE, MM5-EURAD, and CALIOPE. They each operate daily using different horizontal resolutions (10 km × 10 km, 5 km × 5 km, and 4 km × 4 km, respectively), specific physical and chemical parameterizations, and their own emission pre-processors (with a common EMEP emission database source but different spatial disaggregation methodologies). The operational BSC-DREAM8b model is coupled offline within the aforementioned air quality systems to provide the Saharan dust contribution to particulate matter. Bias-correction studies have demonstrated the benefit of using past observational data to reduce systematic model forecast errors. The present contribution aims to evaluate the application of two bias-correction techniques, the multiplicative ratio and the Kalman filter, in order to improve air quality forecasts for Portugal. Both techniques are applied to the three modelling systems over the full year of 2010. Raw and unbiased model results for the main atmospheric pollutants (O3, NO2, SO2, PM10, and PM2.5) are analysed and compared with data from 18 monitoring stations distributed within inland Portugal on an hourly basis. Statistical analysis shows that both bias-correction techniques improve the raw forecast skills (for all the modelling systems and pollutants). In the case of O3 max-8 h, correlation coefficients improve by 19–45%, from 0.56–0.81 (raw models) to 0.78–0.86 (corrected models). PM2.5 also presents significant improvements, for example correlation coefficients increase by more than 50% (with both techniques), reaching values between 0.50 and 0.64. The corrected primary pollutants NO2 and SO2 demonstrate significant relative improvements compared to O3, mostly because the original modelling system skills are lower for those species. Although the applied techniques have different mathematical formulations and complexity levels, there are comparable answers for all of the forecasting systems. Analysis performed over specific situations such as air quality episodes and cases of unvalidated or missing data reveals different behaviours of the bias-correction techniques under study. The results confirm the advantage of the application of bias-correction techniques for air quality forecasts. Both techniques can be applied routinely in operational forecast systems and they will be useful to provide accurate alerts about exceedances to the population.

Application of Partial Least-Squares Regression in the Prediction of Concrete Strength Taking into Account Interactional Items Among Independent Variables

Based on partial least-squares regression taking into account interactional items among independent variables, this paper had a prediction on concrete strength at the 28th day. Taking proportion of flyash in cementing material, usage amount of cementing material, ash-water ratio as independent variables , and concrete strength at the 28th day as dependent variable , the forecast model of concrete strength was obtained. It was found that press residual value decreased with the increase of number of latent variables, and number of latent variables were three according to Press residual value versus number of latent variables. The normal regression coefficient of ash-water ratio was the largest in three influence factors, this indicated that the influence of ash-water ratio was largest to concrete strength at the 28th day; The determination coefficient of forecast model obtained in this paper was 0.9353, the error of forecast model was. The following conclusion can be drawn that, the model is accurate and credible, and the partial least-squares regression taking into account interactional items among independent variables is a eximious non-linear method, and it is worthy to spread its application in the forecast analysis of concrete strength at the 28th day.

Analogue-dynamical prediction of monsoon precipitation in Northeast China based on dynamic and optimal configuration of multiple predictors

Based on the National Climate Center (NCC) of China operational seasonal prediction model results for the period 1983–2009 and the US National Weather Service Climate Prediction Center merged analysis of precipitation in the same period, together with the 74 circulation indices of NCC Climate System Diagnostic Division and 40 climate indices of NOAA of US during 1951–2009, an analogue-dynamical technique for objective and quantitative prediction of monsoon precipitation in Northeast China is proposed and implemented. Useful information is extracted from the historical data to estimate the model forecast errors. Dominant predictors and the predictors that exhibit evolving analogues are identified through cross validating the anomaly correlation coefficients (ACC) among single predictors, meanwhile with reference of the results from the dynamic analogue bias correction using four analogue samples. Next, an optimal configuration of multiple predictors is set up and compared with historical optimal multi-predictor configurations and then dynamically adjusted. Finally, the model errors are evaluated and utilized to correct the NCC operational seasonal prediction model results, and the forecast of monsoon precipitation is obtained at last. The independent sample validation shows that this technique has effectively improved the monsoon precipitation prediction skill during 2005–2009. This study demonstrates that the analogue-dynamical approach is feasible in operational prediction of monsoon precipitation.

Observational Error Estimation of FORMOSAT-3/COSMIC GPS Radio Occultation Data

Abstract The Global Positioning System (GPS) radio occultation (RO) technique is becoming a robust global observing system. GPS RO refractivity is typically modeled at the ray perigee point by a “local refractivity operator” in a data assimilation system. Such modeling does not take into account the horizontal gradients that affect the GPS RO refractivity. A new observable (linear excess phase), defined as an integral of the refractivity along some fixed ray path within the model domain, has been developed in earlier studies to account for the effect of horizontal gradients. In this study, the error statistics of both observables (refractivity and linear excess phase) are estimated using the GPS RO data from the Formosa Satellite 3–Constellation Observing System for Meteorology, Ionosphere and Climate (FORMOSAT-3/COSMIC) mission. The National Meteorological Center (NMC) method, which is based on lagged forecast differences, is applied for evaluation of the model forecast errors that are used for estimation of the GPS RO observational errors. Also used are Weather Research and Forecasting (WRF) model forecasts in the East Asia region at 45-km resolution for one winter month (mid-January to mid-February) and one summer month (mid-August to mid-September) in 2007. Fractional standard deviations of the observational errors of refractivity and linear excess phase both show an approximately linear decrease with height in the troposphere and a slight increase above the tropopause; their maximum magnitude is about 2.2% (2.5%) for refractivity and 1.1% (1.3%) for linear excess phase in the lowest 2 km for the winter (summer) month. An increase of both fractional observational errors near the surface in the summer month is attributed mainly to a larger amount of water vapor. The results indicate that the fractional observational error of refractivity is about twice as large as that of linear excess phase, regardless of season. The observational errors of both linear excess phase and refractivity are much less latitude dependent for summer than for winter. This difference is attributed to larger latitudinal variations of the specific humidity in winter.