Assimilation of radar-based nowcast into HIRLAM NWP model

David Getreuer Jensen, Claus Petersen, Michael R. Rasmussen

Research output: Contribution to journalJournal articleResearchpeer-review

5 Citations (Scopus)

Abstract

The present study introduces a nowcast scheme that assimilates radar extrapolation data (RED) into a nowcasting version of the high resolution limited area model (HIRLAM) numerical weather prediction (NWP) model covering the area of Denmark. The RED are based on the Co-TREC (tracking radar echoes by correlation) methodology and are generated from cleaned radar mosaics from the Danish weather radar network. The assimilation technique is a newly developed method that increases model precipitation by increasing low-level convergence and decreasing convergence aloft in order to increase the vertical velocity. A selection criterion based on spatial coverage and temporal duration is applied to identify rain events for evaluation. Eight events from August 2010 comply with the criteria and additionally, an extreme event on 2 July 2011 is also evaluated. The RED assimilated NWP model is verified against a non-assimilated NWP reference run using the fractional skill score. In all cases, the predictability of the RED assimilated NWP model has been increased compared with the non-RED assimilated NWP model. After ending the RED assimilation, the positive effects continue until the skill converges towards the reference run. The level of improved predictability relies on the RED quality, which again relies on the type of event.
Original languageEnglish
JournalMeteorological Applications
Volume22
Issue number3
Pages (from-to)485–494
Number of pages10
ISSN1350-4827
DOIs
Publication statusPublished - 2015

Fingerprint

radar
weather
prediction
assimilation
nowcasting
extreme event
data quality
data assimilation
methodology

Keywords

  • NWP
  • Radar
  • Nowcast
  • Assimilation
  • Radar extrapolation nowcast
  • Fractional skill score
  • FSS

Cite this

Jensen, David Getreuer ; Petersen, Claus ; Rasmussen, Michael R. / Assimilation of radar-based nowcast into HIRLAM NWP model. In: Meteorological Applications. 2015 ; Vol. 22, No. 3. pp. 485–494.
@article{6a607d2f864444eab1f9e407acaef08b,
title = "Assimilation of radar-based nowcast into HIRLAM NWP model",
abstract = "The present study introduces a nowcast scheme that assimilates radar extrapolation data (RED) into a nowcasting version of the high resolution limited area model (HIRLAM) numerical weather prediction (NWP) model covering the area of Denmark. The RED are based on the Co-TREC (tracking radar echoes by correlation) methodology and are generated from cleaned radar mosaics from the Danish weather radar network. The assimilation technique is a newly developed method that increases model precipitation by increasing low-level convergence and decreasing convergence aloft in order to increase the vertical velocity. A selection criterion based on spatial coverage and temporal duration is applied to identify rain events for evaluation. Eight events from August 2010 comply with the criteria and additionally, an extreme event on 2 July 2011 is also evaluated. The RED assimilated NWP model is verified against a non-assimilated NWP reference run using the fractional skill score. In all cases, the predictability of the RED assimilated NWP model has been increased compared with the non-RED assimilated NWP model. After ending the RED assimilation, the positive effects continue until the skill converges towards the reference run. The level of improved predictability relies on the RED quality, which again relies on the type of event.",
keywords = "NWP, Radar, Nowcast, Assimilation, Radar extrapolation nowcast, Fractional skill score, FSS, NWP, Radar, Nowcast, Assimilation, Radar extrapolation nowcast, Fractional skill score, FSS",
author = "Jensen, {David Getreuer} and Claus Petersen and Rasmussen, {Michael R.}",
year = "2015",
doi = "10.1002/met.1479",
language = "English",
volume = "22",
pages = "485–494",
journal = "Meteorological Applications",
issn = "1350-4827",
publisher = "Wiley",
number = "3",

}

Assimilation of radar-based nowcast into HIRLAM NWP model. / Jensen, David Getreuer; Petersen, Claus; Rasmussen, Michael R.

In: Meteorological Applications, Vol. 22, No. 3, 2015, p. 485–494.

Research output: Contribution to journalJournal articleResearchpeer-review

TY - JOUR

T1 - Assimilation of radar-based nowcast into HIRLAM NWP model

AU - Jensen, David Getreuer

AU - Petersen, Claus

AU - Rasmussen, Michael R.

PY - 2015

Y1 - 2015

N2 - The present study introduces a nowcast scheme that assimilates radar extrapolation data (RED) into a nowcasting version of the high resolution limited area model (HIRLAM) numerical weather prediction (NWP) model covering the area of Denmark. The RED are based on the Co-TREC (tracking radar echoes by correlation) methodology and are generated from cleaned radar mosaics from the Danish weather radar network. The assimilation technique is a newly developed method that increases model precipitation by increasing low-level convergence and decreasing convergence aloft in order to increase the vertical velocity. A selection criterion based on spatial coverage and temporal duration is applied to identify rain events for evaluation. Eight events from August 2010 comply with the criteria and additionally, an extreme event on 2 July 2011 is also evaluated. The RED assimilated NWP model is verified against a non-assimilated NWP reference run using the fractional skill score. In all cases, the predictability of the RED assimilated NWP model has been increased compared with the non-RED assimilated NWP model. After ending the RED assimilation, the positive effects continue until the skill converges towards the reference run. The level of improved predictability relies on the RED quality, which again relies on the type of event.

AB - The present study introduces a nowcast scheme that assimilates radar extrapolation data (RED) into a nowcasting version of the high resolution limited area model (HIRLAM) numerical weather prediction (NWP) model covering the area of Denmark. The RED are based on the Co-TREC (tracking radar echoes by correlation) methodology and are generated from cleaned radar mosaics from the Danish weather radar network. The assimilation technique is a newly developed method that increases model precipitation by increasing low-level convergence and decreasing convergence aloft in order to increase the vertical velocity. A selection criterion based on spatial coverage and temporal duration is applied to identify rain events for evaluation. Eight events from August 2010 comply with the criteria and additionally, an extreme event on 2 July 2011 is also evaluated. The RED assimilated NWP model is verified against a non-assimilated NWP reference run using the fractional skill score. In all cases, the predictability of the RED assimilated NWP model has been increased compared with the non-RED assimilated NWP model. After ending the RED assimilation, the positive effects continue until the skill converges towards the reference run. The level of improved predictability relies on the RED quality, which again relies on the type of event.

KW - NWP

KW - Radar

KW - Nowcast

KW - Assimilation

KW - Radar extrapolation nowcast

KW - Fractional skill score

KW - FSS

KW - NWP

KW - Radar

KW - Nowcast

KW - Assimilation

KW - Radar extrapolation nowcast

KW - Fractional skill score

KW - FSS

U2 - 10.1002/met.1479

DO - 10.1002/met.1479

M3 - Journal article

VL - 22

SP - 485

EP - 494

JO - Meteorological Applications

JF - Meteorological Applications

SN - 1350-4827

IS - 3

ER -