Monitoring of Observation and Analysis Quality by a Data Assimilation System

Hollingsworth, A.; Shaw, David B.; Lönnberg, P.; Illari, Lodovica; Arpe, Klaus; Simmons, A. J.

doi:10.1175/1520-0493(1986)114<0861:mooaaq>2.0.co;2

Cited by 112 publications

(57 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…For ease of reference, the lengths of the various increments are listed in Table 1. We thus verify the first part of the criterion of Hollingsworth et al (1986) for a successful data assimilation cycle which states that the forecast increment, here of length 4.81 m s −1 , should be larger than the analysis increment, here of length 2.65 m s −1 . It would appear that the length of the model error, here 1.39 m s −1 , is fairly significant compared to the length of the total forecast, here 4.81 m s −1 .…”

Section: Length Of Incrementssupporting

confidence: 63%

“…We measured the magnitude of the forecast component F , the analysis component A, and, if initialization is performed, of the initialization component I . The relative sizes of these components were found to be ordered in accordance with the Hollingsworth et al (1986) criteria for a properly functioning data assimilation system, i.e. F > A > I.…”

Section: Discussionmentioning

confidence: 80%

“…Similarly we obtain a value of 2.40 m s −1 for the difference between the state at 6 and 9 h. The approximation that the observations are valid at the central analysis time may have been appropriate when radiosonde and other synoptic observations were predominant, but given the ever-increasing amount of satellite and other asynoptic data, is equivalent to assuming negligible model evolution in the 6 h window. Such an approximation is clearly severe and is, in fact, inconsistent with the earlier condition, from Hollingsworth et al (1986), that the forecast increment must be bigger than the analysis increment.…”

Section: Length Of Incrementsmentioning

confidence: 77%

“…Its value of 0.35 m s −1 is well below the 2.65 m s −1 of the analysis increment. Thus our EnKF respects the second part of the criterion of Hollingsworth et al (1986) which states that the analysis increment should be bigger than the initialization increment.…”

Section: Length Of Incrementsmentioning

confidence: 99%

See 3 more Smart Citations

Ensemble Kalman filtering

Houtekamer

Mitchell

2005

Quart J Royal Meteoro Soc

338

203

View full text Add to dashboard Cite

SUMMARYAn ensemble Kalman filter (EnKF) has been implemented at the Canadian Meteorological Centre to provide an ensemble of initial conditions for the medium-range ensemble prediction system. This demonstrates that the EnKF can be used for operational atmospheric data assimilation.We show how the EnKF relates to the Kalman filter. In particular, to make the ensemble approximation feasible, we have to use a fairly small ensemble with many less members than either the number of model coordinates, or the number of independent observations, or the (unknown) dimension of the dynamical system. To nevertheless obtain good results, we must (i) counter the tendency of the ensemble spread to underestimate the true error, and (ii) localize the ensemble covariances. The localization is severe and leads to imbalance in the initial conditions.The operational EnKF is used to investigate to what extent our system respects the underlying hypotheses of both the Kalman filter and its ensemble approximation. In particular, we quantify the imbalance in the initial conditions and the magnitude of the model-error component. The occurrence of imbalance constrains the ways in which time interpolation can be performed and in which parametrized model error can be added. With this study we hope to obtain and provide guidance for further improvements to the EnKF.

show abstract

Section: Length Of Incrementssupporting

confidence: 63%

Section: Discussionmentioning

confidence: 80%

Section: Length Of Incrementsmentioning

confidence: 77%

Section: Length Of Incrementsmentioning

confidence: 99%

See 2 more Smart Citations

Ensemble Kalman filtering

Houtekamer

Mitchell

2005

Quart J Royal Meteoro Soc

338

203

View full text Add to dashboard Cite

show abstract

“…Large-scale meteorological conditions were described using global gridded analyses prepared by the European Center for Medium-Range Weather Forecasting (ECMWF) [Bengtsson, 1985;Hollingsworth et al, 1986]. These analyses were assimilated from surface observations, radiosonde data, aircraft reports, satellite-derived soundings, and cloud drift winds.…”

Section: Introductionmentioning

confidence: 99%

Influence of a middle‐latitude cyclone on tropospheric ozone distributions during a period of TRACE A

Loring¹,

Fuelberg²,

Fishman³

et al. 1996

J. Geophys. Res.

View full text Add to dashboard Cite

Abstract. A middle-latitude cyclone occurring during the Transport and AtmosphericChemistry near the Equator-Atlantic (TRACE A) experiment is examined to determine its influence on distributions of tropospheric ozone over the South Atlantic Ocean. A maximum of tropospheric ozone is located in the vicinity of this cyclone on October 3, 1992. Flight level data and meteorological analyses indicate a downward protrusion of dry, ozone-rich stratospheric air near the cyclone, i.e., a tropopause fold. Backward trajectories show that air parcels arriving in the upper troposphere of the cyclone originate in the stratosphere. Forward trajectories are calculated from these locations having stratospheric histories. They indicate that some air is transported as far north as 22øS, subsiding into the middle troposphere along the southern fringes of a region of enhanced tropospheric ozone that is located west of Africa on October 6. Backward trajectories then are computed along the Greenwich meridian over much of the South Atlantic Ocean. This axis passes through the tropospheric ozone maximum west of Africa and the region of strong horizontal ozone gradients along its southern border. Results indicate that most air parcels arriving north of 20øS (in the ozone-rich region) originate over Africa. Conversely, most parcels arriving south of 20øS (where there is less ozone) originate from the west, passing over the southern half of South America. Thus the tropospheric ozone maximum west of Africa on October 6 appears to be attributable to outflow from Africa, with stratospheric transport being much less important. Formerly stratospheric air near the cyclone on October 3 also is transported forward into the middle troposphere near Madagascar where there is a second maximum of tropospheric ozone on October 6. Backward trajectories from this region indicate that middle-latitude systems exert a much greater influence here than over the South Atlantic. This area experiences relatively little outflow from Africa during our period of study. [1991] used data from ozonesondes to conclude that only 25% of the ozone mixing ratio at 300 mbar for Payerne, Switzerland, was due to stratospheric intrusions. The remainder was attributed to in situ photochemical production. Since tropospheric ozone is both transported from the stratosphere and produced photochemically [Levy, 1988], it is important to understand their relative roles in producing the ozone anomaly over the South Atlantic Ocean.

show abstract

Historical Overview of Numerical Weather Prediction

Kalnay

2003

Handbook of Weather, Climate, and Water

View full text Add to dashboard Cite

Monitoring of Observation and Analysis Quality by a Data Assimilation System

Cited by 112 publications

References 0 publications

Ensemble Kalman filtering

Ensemble Kalman filtering

Influence of a middle‐latitude cyclone on tropospheric ozone distributions during a period of TRACE A

Historical Overview of Numerical Weather Prediction

Contact Info

Product

Resources

About