Search Results
You are looking at 1 - 10 of 17 items for
- Author or Editor: Linus Magnusson x
- Refine by Access: All Content x
Abstract
Accurate estimation of the position of the departure points (d.p.) is crucial for the accuracy of a semi-Lagrangian NWP model. This calculation is often performed applying an implicit discretization to a kinematic equation solved by a fixed-point iteration scheme. A small number of iterations is typically used, assuming that this is sufficient for convergence. This assumption, derived from a past theoretical analysis, is revisited here. Analyzing the convergence of a generic d.p. iteration scheme and testing the ECMWF Integrated Forecast System (IFS) model, it is demonstrated that 2–3 iterations may not be sufficient for convergence to satisfactory accuracy in a modern high-resolution global model. Large forecast improvements can be seen by increasing the number of iterations. The extratropical geopotential error decreases and the simulated vertical structure of tropical cyclones improves. These findings prompted the implementation of an algorithm in which stopping criteria based on estimated convergence rates are used to “dynamically” stop d.p. iterations when an error tolerance criterion is satisfied. This is applied consistently to the nonlinear forecast, tangent linear, and adjoint models used by the ECMWF data assimilation system (4DVAR). Although the additional benefit of dynamic iteration is small, its testing reinforces the conclusion that a larger number of iterations is needed in regions of strong winds and shear. Furthermore, experiments suggest that dynamic iteration may prevent occasional 4DVAR failures in cases of strong stratospheric cross-polar flow in which the tangent linear model becomes unstable.
Abstract
Accurate estimation of the position of the departure points (d.p.) is crucial for the accuracy of a semi-Lagrangian NWP model. This calculation is often performed applying an implicit discretization to a kinematic equation solved by a fixed-point iteration scheme. A small number of iterations is typically used, assuming that this is sufficient for convergence. This assumption, derived from a past theoretical analysis, is revisited here. Analyzing the convergence of a generic d.p. iteration scheme and testing the ECMWF Integrated Forecast System (IFS) model, it is demonstrated that 2–3 iterations may not be sufficient for convergence to satisfactory accuracy in a modern high-resolution global model. Large forecast improvements can be seen by increasing the number of iterations. The extratropical geopotential error decreases and the simulated vertical structure of tropical cyclones improves. These findings prompted the implementation of an algorithm in which stopping criteria based on estimated convergence rates are used to “dynamically” stop d.p. iterations when an error tolerance criterion is satisfied. This is applied consistently to the nonlinear forecast, tangent linear, and adjoint models used by the ECMWF data assimilation system (4DVAR). Although the additional benefit of dynamic iteration is small, its testing reinforces the conclusion that a larger number of iterations is needed in regions of strong winds and shear. Furthermore, experiments suggest that dynamic iteration may prevent occasional 4DVAR failures in cases of strong stratospheric cross-polar flow in which the tangent linear model becomes unstable.
Abstract
During the past 30 years the skill in ECMWF numerical forecasts has steadily improved. There are three major contributing factors: 1) improvements in the forecast model, 2) improvements in the data assimilation, and 3) the increased number of available observations. In this study the authors are investigating the relative contribution from these three components by using the simple error growth model introduced in a previous study by Lorenz and extended in another study by Dalcher and Kalnay, together with the results from the ECMWF Re-Analysis Interim (ERA-Interim) forecasts where the improvement is only due to an increased number of observations. The authors are also applying the growth model on “lagged” forecast differences in order to investigate the usefulness of the forecast jumpiness as a diagnostic tool for improvements in the forecasts. The main finding is that the main contribution to the reduced forecast error comes from significant initial condition error reductions between 1996 and 2001 together with continuous model improvements. The changes in the available observations contributed to a lesser degree, but the authors note that all the ERA-Interim forecasts are from the satellite era and here the focus is on the midtroposphere in the extratropics. Regarding the jumpiness in the forecasts, this is mainly a function of the error in the initial conditions and is therefore an insufficient tool to investigate improvements in the full forecasting system.
Abstract
During the past 30 years the skill in ECMWF numerical forecasts has steadily improved. There are three major contributing factors: 1) improvements in the forecast model, 2) improvements in the data assimilation, and 3) the increased number of available observations. In this study the authors are investigating the relative contribution from these three components by using the simple error growth model introduced in a previous study by Lorenz and extended in another study by Dalcher and Kalnay, together with the results from the ECMWF Re-Analysis Interim (ERA-Interim) forecasts where the improvement is only due to an increased number of observations. The authors are also applying the growth model on “lagged” forecast differences in order to investigate the usefulness of the forecast jumpiness as a diagnostic tool for improvements in the forecasts. The main finding is that the main contribution to the reduced forecast error comes from significant initial condition error reductions between 1996 and 2001 together with continuous model improvements. The changes in the available observations contributed to a lesser degree, but the authors note that all the ERA-Interim forecasts are from the satellite era and here the focus is on the midtroposphere in the extratropics. Regarding the jumpiness in the forecasts, this is mainly a function of the error in the initial conditions and is therefore an insufficient tool to investigate improvements in the full forecasting system.
Abstract
In this paper a study aimed at comparing the perturbation methodologies based on the singular vector ensemble prediction system (SV-EPS) and the breeding vector ensemble prediction system (BV-EPS) in the same model environment is presented. A simple breeding system (simple BV-EPS) as well as one with regional rescaling dependent on an estimate of the analysis error variance (masked BV-EPS) were used. The ECMWF Integrated Forecast System has been used and the three experiments are compared for 46 forecast cases between 1 December 2005 and 15 January 2006. By studying the distribution of the perturbation energy it was possible to see large differences between the experiments initially, but after 48 h the distributions have converged. Using probabilistic scores, these results show that SV-EPS has a somewhat better performance for the Northern Hemisphere compared to BV-EPS. For the Southern Hemisphere masked BV-EPS and SV-EPS yield almost equal results. For the tropics the masked breeding ensemble shows the best performance during the first 6 days. One reason for this is the current setup of the singular vector ensemble at ECMWF yielding in general very low initial perturbation amplitudes in the tropics.
Abstract
In this paper a study aimed at comparing the perturbation methodologies based on the singular vector ensemble prediction system (SV-EPS) and the breeding vector ensemble prediction system (BV-EPS) in the same model environment is presented. A simple breeding system (simple BV-EPS) as well as one with regional rescaling dependent on an estimate of the analysis error variance (masked BV-EPS) were used. The ECMWF Integrated Forecast System has been used and the three experiments are compared for 46 forecast cases between 1 December 2005 and 15 January 2006. By studying the distribution of the perturbation energy it was possible to see large differences between the experiments initially, but after 48 h the distributions have converged. Using probabilistic scores, these results show that SV-EPS has a somewhat better performance for the Northern Hemisphere compared to BV-EPS. For the Southern Hemisphere masked BV-EPS and SV-EPS yield almost equal results. For the tropics the masked breeding ensemble shows the best performance during the first 6 days. One reason for this is the current setup of the singular vector ensemble at ECMWF yielding in general very low initial perturbation amplitudes in the tropics.
Abstract
Subseasonal forecasts lie between medium-range and seasonal time scales with an emerging attention due to the relevance in society and by the scientific challenges involved. This study aims to (i) evaluate the development of systematic errors with lead time in the European Centre for Medium-Range Weather Forecasts (ECMWF) ensemble forecasts of surface-related variables during late spring and summer, and (ii) investigate potential relationships between the systematic errors and predictive skill. The evaluation is performed over the Northern Hemisphere midlatitudes, focusing on several regions with different climate characteristics. The results indicate five key bias patterns: (i) cold bias of daily maximum temperature (mx2t) in the April–May forecasts at all lead times in most regions; (ii) central North America with a warm bias mostly in the daily minimum temperature (mn2t); (iii) east of the Caspian Sea region with a warm and dry bias; (iv) western and Mediterranean Europe with a cold bias in mn2t mainly in April–May forecasts; and (v) continental Europe with a cold bias in the mx2t and warm bias of mn2t in the June–July forecasts. We also found substantial deviations of soil moisture and terrestrial water storage variation in most regions compared to the fifth-generation ECMWF atmospheric reanalysis (ERA5). Despite the large differences in the systematic error characteristics among the different regions, there is little relation to the skill of the subseasonal forecasts. The systematic temperature biases require further attention from model developers as diurnal cycle improvements could enhance some of the potential predictability coming from the long-memory effect of soil moisture.
Abstract
Subseasonal forecasts lie between medium-range and seasonal time scales with an emerging attention due to the relevance in society and by the scientific challenges involved. This study aims to (i) evaluate the development of systematic errors with lead time in the European Centre for Medium-Range Weather Forecasts (ECMWF) ensemble forecasts of surface-related variables during late spring and summer, and (ii) investigate potential relationships between the systematic errors and predictive skill. The evaluation is performed over the Northern Hemisphere midlatitudes, focusing on several regions with different climate characteristics. The results indicate five key bias patterns: (i) cold bias of daily maximum temperature (mx2t) in the April–May forecasts at all lead times in most regions; (ii) central North America with a warm bias mostly in the daily minimum temperature (mn2t); (iii) east of the Caspian Sea region with a warm and dry bias; (iv) western and Mediterranean Europe with a cold bias in mn2t mainly in April–May forecasts; and (v) continental Europe with a cold bias in the mx2t and warm bias of mn2t in the June–July forecasts. We also found substantial deviations of soil moisture and terrestrial water storage variation in most regions compared to the fifth-generation ECMWF atmospheric reanalysis (ERA5). Despite the large differences in the systematic error characteristics among the different regions, there is little relation to the skill of the subseasonal forecasts. The systematic temperature biases require further attention from model developers as diurnal cycle improvements could enhance some of the potential predictability coming from the long-memory effect of soil moisture.
Abstract
One desirable property within an ensemble forecast system is to have a one-to-one ratio between the root-mean-square error (rmse) of the ensemble mean and the standard deviation of the ensemble (spread). The ensemble spread and forecast error within the ECMWF ensemble prediction system has been extrapolated beyond 10 forecast days using a simple model for error growth. The behavior of the ensemble spread and the rmse at the time of the deterministic predictability are compared with derived relations of rmse at the infinite forecast length and the characteristic variability of the atmosphere in the limit of deterministic predictability. Utilizing this methodology suggests that the forecast model and the atmosphere do not have the same variability, which raises the question of how to obtain a perfect ensemble.
Abstract
One desirable property within an ensemble forecast system is to have a one-to-one ratio between the root-mean-square error (rmse) of the ensemble mean and the standard deviation of the ensemble (spread). The ensemble spread and forecast error within the ECMWF ensemble prediction system has been extrapolated beyond 10 forecast days using a simple model for error growth. The behavior of the ensemble spread and the rmse at the time of the deterministic predictability are compared with derived relations of rmse at the infinite forecast length and the characteristic variability of the atmosphere in the limit of deterministic predictability. Utilizing this methodology suggests that the forecast model and the atmosphere do not have the same variability, which raises the question of how to obtain a perfect ensemble.
Abstract
The Thames basin experienced 12 major Atlantic depressions in winter 2013/14, leading to extensive and prolonged fluvial and groundwater flooding. This exceptional weather coincided with highly anomalous meteorological conditions across the globe. Atmospheric relaxation experiments, whereby conditions within specified regions are relaxed toward a reanalysis, have been used to investigate teleconnection patterns. However, no studies have examined whether improvements to seasonal meteorological forecasts translate into more skillful seasonal hydrological forecasts. This study applied relaxation experiments to reforecast the 2013/14 floods for three Thames basin catchments with different hydrogeological characteristics. The tropics played an important role in the development of extreme conditions over the Thames basin. The greatest hydrological forecasting skill was associated with the tropical Atlantic and less with the tropical Pacific, although both captured seasonal meteorological flow anomalies. Relaxation applied over the northeastern Atlantic produced confident ensemble forecasts, but hydrological extremes were underpredicted; this was unexpected with relaxation applied so close to the United Kingdom. Streamflow was most skillfully forecast for the catchment representing a large drainage area with high peak flow. Permeable lithology and antecedent conditions were important for skillfully forecasting groundwater levels. Atmospheric relaxation experiments can improve our understanding of extratropical anomalies and the potential predictability of extreme events such as the Thames 2013/14 floods. Seasonal hydrological forecasts differed from what was expected from the meteorology alone, and thus knowledge is gained by considering both components. In the densely populated Thames basin, considering the local hydrogeological context can provide an effective early alert of potential high-impact events, allowing for better preparedness.
Abstract
The Thames basin experienced 12 major Atlantic depressions in winter 2013/14, leading to extensive and prolonged fluvial and groundwater flooding. This exceptional weather coincided with highly anomalous meteorological conditions across the globe. Atmospheric relaxation experiments, whereby conditions within specified regions are relaxed toward a reanalysis, have been used to investigate teleconnection patterns. However, no studies have examined whether improvements to seasonal meteorological forecasts translate into more skillful seasonal hydrological forecasts. This study applied relaxation experiments to reforecast the 2013/14 floods for three Thames basin catchments with different hydrogeological characteristics. The tropics played an important role in the development of extreme conditions over the Thames basin. The greatest hydrological forecasting skill was associated with the tropical Atlantic and less with the tropical Pacific, although both captured seasonal meteorological flow anomalies. Relaxation applied over the northeastern Atlantic produced confident ensemble forecasts, but hydrological extremes were underpredicted; this was unexpected with relaxation applied so close to the United Kingdom. Streamflow was most skillfully forecast for the catchment representing a large drainage area with high peak flow. Permeable lithology and antecedent conditions were important for skillfully forecasting groundwater levels. Atmospheric relaxation experiments can improve our understanding of extratropical anomalies and the potential predictability of extreme events such as the Thames 2013/14 floods. Seasonal hydrological forecasts differed from what was expected from the meteorology alone, and thus knowledge is gained by considering both components. In the densely populated Thames basin, considering the local hydrogeological context can provide an effective early alert of potential high-impact events, allowing for better preparedness.
Abstract
A method has been developed to forecast seasonal landfall risk using ensembles of cyclone tracks generated by ECMWF’s seasonal forecast system 4. The method has been applied to analyze and retrospectively forecast the landfall risk along the North American coast. The main result is that the method can be used to forecast landfall for some parts of the coast, but the skill is lower than for basinwide forecasts of activity. The rank correlations between forecasts issued on 1 May and observations are 0.6 for basinwide tropical cyclone number and 0.5 for landfall anywhere along the coast. When the forecast period is limited to the peak of the hurricane season, the landfall correlation increases to 0.6. Moreover, when the forecast issue date is pushed forward to 1 August, basinwide tropical cyclone and hurricane correlations increase to 0.7 and 0.8, respectively, whereas landfall correlations improve less. The quality of the forecasts is in line with that obtained by others.
Abstract
A method has been developed to forecast seasonal landfall risk using ensembles of cyclone tracks generated by ECMWF’s seasonal forecast system 4. The method has been applied to analyze and retrospectively forecast the landfall risk along the North American coast. The main result is that the method can be used to forecast landfall for some parts of the coast, but the skill is lower than for basinwide forecasts of activity. The rank correlations between forecasts issued on 1 May and observations are 0.6 for basinwide tropical cyclone number and 0.5 for landfall anywhere along the coast. When the forecast period is limited to the peak of the hurricane season, the landfall correlation increases to 0.6. Moreover, when the forecast issue date is pushed forward to 1 August, basinwide tropical cyclone and hurricane correlations increase to 0.7 and 0.8, respectively, whereas landfall correlations improve less. The quality of the forecasts is in line with that obtained by others.
Abstract
On 30 October 2012 Hurricane Sandy made landfall on the U.S. East Coast with a devastating impact. Here the performance of the ECMWF forecasts (both high resolution and ensemble) are evaluated together with ensemble forecasts from other numerical weather prediction centers, available from The Observing System Research and Predictability Experiment (THORPEX) Interactive Grand Global Ensemble (TIGGE) archive. The sensitivity to sea surface temperature (SST) and model resolution for the ECMWF forecasts are explored. The results show that the ECMWF forecasts provided a clear indication of the landfall from 7 days in advance. Comparing ensemble forecasts from different centers, the authors find the ensemble forecasts from ECMWF to be the most consistent in the forecast of the landfall of Sandy on the New Jersey coastline. The impact of the warm SST anomaly off the U.S. East Coast is investigated by running sensitivity experiments with climatological SST instead of persisting the SST anomaly from the analysis. The results show that the SST anomaly had a small effect on Sandy’s track in the forecast, but the forecasts initialized with the warm SST anomaly feature a more intense system in terms of the depth of the cyclone, wind speeds, and precipitation. Furthermore, the role of spatial resolution is investigated by comparing four global simulations, spanning from TL159 (150 km) to TL3999 (5 km) horizontal resolution. Forecasts from 3 and 5 days before the landfall are evaluated. While all resolutions predict Sandy’s landfall, at very high resolution the tropical cyclone intensity and the oceanic wave forecasts are greatly improved.
Abstract
On 30 October 2012 Hurricane Sandy made landfall on the U.S. East Coast with a devastating impact. Here the performance of the ECMWF forecasts (both high resolution and ensemble) are evaluated together with ensemble forecasts from other numerical weather prediction centers, available from The Observing System Research and Predictability Experiment (THORPEX) Interactive Grand Global Ensemble (TIGGE) archive. The sensitivity to sea surface temperature (SST) and model resolution for the ECMWF forecasts are explored. The results show that the ECMWF forecasts provided a clear indication of the landfall from 7 days in advance. Comparing ensemble forecasts from different centers, the authors find the ensemble forecasts from ECMWF to be the most consistent in the forecast of the landfall of Sandy on the New Jersey coastline. The impact of the warm SST anomaly off the U.S. East Coast is investigated by running sensitivity experiments with climatological SST instead of persisting the SST anomaly from the analysis. The results show that the SST anomaly had a small effect on Sandy’s track in the forecast, but the forecasts initialized with the warm SST anomaly feature a more intense system in terms of the depth of the cyclone, wind speeds, and precipitation. Furthermore, the role of spatial resolution is investigated by comparing four global simulations, spanning from TL159 (150 km) to TL3999 (5 km) horizontal resolution. Forecasts from 3 and 5 days before the landfall are evaluated. While all resolutions predict Sandy’s landfall, at very high resolution the tropical cyclone intensity and the oceanic wave forecasts are greatly improved.
Abstract
In their comment, Žagar and Szunyogh raised concerns about a recent study by Zhang et al. that examined the predictability limit of midlatitude weather using two up-to-date global models. Zhang et al. showed that deterministic weather forecast may, at best, be extended by 5 days, assuming we could achieve minimal initial-condition uncertainty (e.g., 10% of current operational value) with a nearly perfect model. Žagar and Szunyogh questioned the methodology and the experiments of Zhang et al. Specifically, Žagar and Szunyogh raised issues regarding the effects of model error on the growth of the forecast uncertainty. They also suggested that estimates of the predictability limit could be obtained using a simple parametric model. This reply clarifies the misunderstandings in Žagar and Szunyogh and demonstrates that experiments conducted by Zhang et al. are reasonable. In our view, the model error concern in Žagar and Szunyogh does not apply to the intrinsic predictability limit, which is the key focus of Zhang et al. and the simple parametric model described in Žagar and Szunyogh does not serve the purpose of Zhang et al.
Abstract
In their comment, Žagar and Szunyogh raised concerns about a recent study by Zhang et al. that examined the predictability limit of midlatitude weather using two up-to-date global models. Zhang et al. showed that deterministic weather forecast may, at best, be extended by 5 days, assuming we could achieve minimal initial-condition uncertainty (e.g., 10% of current operational value) with a nearly perfect model. Žagar and Szunyogh questioned the methodology and the experiments of Zhang et al. Specifically, Žagar and Szunyogh raised issues regarding the effects of model error on the growth of the forecast uncertainty. They also suggested that estimates of the predictability limit could be obtained using a simple parametric model. This reply clarifies the misunderstandings in Žagar and Szunyogh and demonstrates that experiments conducted by Zhang et al. are reasonable. In our view, the model error concern in Žagar and Szunyogh does not apply to the intrinsic predictability limit, which is the key focus of Zhang et al. and the simple parametric model described in Žagar and Szunyogh does not serve the purpose of Zhang et al.
Abstract
The Indian Ocean stands out as the region where the state-of-the-art decadal climate predictions of sea surface temperature (SST) perform the best worldwide for forecast times ranging from the second to the ninth year, according to correlation and root-mean-square error (RMSE) scores. This paper investigates the reasons for this high skill by assessing the contributions from the initial conditions, greenhouse gases, solar activity, and volcanic aerosols. The comparison between the SST correlation skill in uninitialized historical simulations and hindcasts initialized from estimates of the observed climate state shows that the high Indian Ocean skill is largely explained by the varying radiative forcings, the latter finding being supported by a set of additional sensitivity experiments. The long-term warming trend is the primary contributor to the high skill, though not the only one. Volcanic aerosols bring additional skill in this region as shown by the comparison between initialized hindcasts taking into account or not the effect of volcanic stratospheric aerosols and by the drop in skill when filtering out their effect in hindcasts that take them into account. Indeed, the Indian Ocean is shown to be the region where the ratio of the internally generated over the externally forced variability is the lowest, where the amplitude of the internal variability has been estimated by removing the effect of long-term warming trend and volcanic aerosols by a multiple least squares linear regression on observed SSTs.
Abstract
The Indian Ocean stands out as the region where the state-of-the-art decadal climate predictions of sea surface temperature (SST) perform the best worldwide for forecast times ranging from the second to the ninth year, according to correlation and root-mean-square error (RMSE) scores. This paper investigates the reasons for this high skill by assessing the contributions from the initial conditions, greenhouse gases, solar activity, and volcanic aerosols. The comparison between the SST correlation skill in uninitialized historical simulations and hindcasts initialized from estimates of the observed climate state shows that the high Indian Ocean skill is largely explained by the varying radiative forcings, the latter finding being supported by a set of additional sensitivity experiments. The long-term warming trend is the primary contributor to the high skill, though not the only one. Volcanic aerosols bring additional skill in this region as shown by the comparison between initialized hindcasts taking into account or not the effect of volcanic stratospheric aerosols and by the drop in skill when filtering out their effect in hindcasts that take them into account. Indeed, the Indian Ocean is shown to be the region where the ratio of the internally generated over the externally forced variability is the lowest, where the amplitude of the internal variability has been estimated by removing the effect of long-term warming trend and volcanic aerosols by a multiple least squares linear regression on observed SSTs.
