Close

1. Identity statement
Reference TypeConference Paper (Conference Proceedings)
Sitemtc-m16b.sid.inpe.br
Holder Codeisadg {BR SPINPE} ibi 8JMKD3MGPCW/3DT298S
Repositorycptec.inpe.br/adm_conf/2005/10.31.16.36
Last Update2006:04.16.15.39.30 (UTC) administrator
Metadata Repositorycptec.inpe.br/adm_conf/2005/10.31.16.36.46
Metadata Last Update2018:06.05.03.42.55 (UTC) administrator
Citation KeyLeiteDiasRotu:2006:EfPrFo
TitleThe effect of precipitaton forecasts and monitoring on hourly hydrological statistical prediction models
FormatCD-ROM, On-line.
Year2006
Access Date2024, Dec. 26
Secondary TypePRE CI
Number of Files1
Size306 KiB
2. Context
Author1 Leite, Eduardo Alvim
2 Dias, Nelson Luís
3 Rotunno Filho, Otto Corrêa
Affiliation1 Instituto Tecnológico SIMEPAR
2 Lemma-Universidade Federal do Paraná
3 COPPE-Universidade Federal do Rio de Janeiro
Author e-Mail Address1 alvim@simepar.br
2 nldias@ufpr.br
3 otto@hidro.ufrj.br
EditorVera, Carolina
Nobre, Carlos
e-Mail Addressalvim@simepar.br
Conference NameInternational Conference on Southern Hemisphere Meteorology and Oceanography, 8 (ICSHMO).
Conference LocationFoz do Iguaçu
Date24-28 Apr. 2006
PublisherAmerican Meteorological Society (AMS)
Publisher City45 Beacon Hill Road, Boston, MA, USA
Pages781-785
Book TitleProceedings
Tertiary TypeOral
OrganizationAmerican Meteorological Society (AMS)
History (UTC)2005-10-31 16:36:46 :: alvim@simepar.br -> adm_conf ::
2005-12-16 01:55:26 :: adm_conf -> alvim@simepar.br ::
2006-03-14 19:23:57 :: alvim@simepar.br -> administrator ::
2006-04-18 21:08:09 :: administrator -> lise@dpi.inpe.br ::
2010-12-28 12:36:36 :: lise@dpi.inpe.br -> administrator ::
2010-12-29 15:57:24 :: administrator -> lise@dpi.inpe.br :: 2006
2010-12-29 16:05:59 :: lise@dpi.inpe.br -> administrator :: 2006
2010-12-29 18:53:10 :: administrator -> banon :: 2006
2011-01-02 17:14:59 :: banon -> administrator :: 2006
2018-06-05 03:42:55 :: administrator -> :: 2006
3. Content and structure
Is the master or a copy?is the master
Content Stagecompleted
Transferable1
KeywordsPrecipitation forecasts
hydrological forecasts
statistical models
AbstractDaily and hourly streamflow forecasts are an important tool for the management of Foz do Areia reservoir in the Iguassu River. Foz do Areia is a hydropower reservoir; it is the first along the Igaussu cascade and also the largest, being responsible for regulating all downstream reservoirs. Historically, streamflow forecasts have relied on a conventional network of streamgages, read twice a day. Ever since 1998, an automated network with GOES telemetry has been implemented, so that now hourly precipitation and river flows are available every 3 hours to be used in streamflow forecasts. We have evaluated the performance of the following modes of hourly streamflow forecasts: a) using only daily observed streamflows; b) using only hourly observed streamflows and c) using both hourly observed streamflows and precipitation, and precipitation forecasts. The streamflow forecasts are based on standard ARIMA models and Kalman filters. These are easily implemented and commonly employed by the Brazilian hydropower companies. They also allow a straightforward assessment of two different aspects of streamflow forecasting: (1) the impact of an automated, telemetric network on the quality of the hourly forecasts and (2) the impact of incorporating precipitation forecasts in the streamflow forecasting. Our main results are the following: when daily observed streamflows are used to predict future hourly streamflows by means of standard ARIMA models, the rising limb of the hydrographs are consistently under-predicted, due to the well-known phase difference in these models. After the observed peak, ARIMA models tend to overshoot and over-predict until a recession is well established. When the hourly data from the automated network are used, it is possible the possible to reduce both kinds of error, which however cannot be completely eliminated, leaving ample room for improvement. Even the application of Kalman filters with streamflow data alone is not able to substatianlly change the picture. Hydrologically this can be explained by the fact that steep catchments which contribute directly to the inflow to Foz do Areia have a relatively short response time, so that precipitation occuring very close to this reservoir is key to good hydrological forecasts. Therefore, when in mode (c) we included simulations of predicted rain with varying accuracy levels, the quality of the streamflow forecasts improved substantially. This was improved by applying a dual-mode, dry/wet Kalman filter, based on the knowledge that precipitation started to occur over the catchment. We were thus able to establish the relationship between precipitation forecast accuracy to streamflow forecast accuracy. It turns out that simply achieving a good forecast of the precipitation class (in our case, we simulated 5 classes) is already enough for most hydrological purposes in 24-hour ahead forecasts.
AreaMET
TypeHydrological variability and modeling
doc Directory Contentaccess
source Directory Content
The_effect_of_precipitation.doc 14/03/2006 16:23 197.0 KiB 
agreement Directory Contentthere are no files
4. Conditions of access and use
data URLhttp://urlib.net/ibi/cptec.inpe.br/adm_conf/2005/10.31.16.36
zipped data URLhttp://urlib.net/zip/cptec.inpe.br/adm_conf/2005/10.31.16.36
Languageen
Target File781-785.pdf
User Groupalvim@simepar.br
administrator
Visibilityshown
5. Allied materials
Host Collectioncptec.inpe.br/nobre/2005/06.02.21.14
cptec.inpe.br/walmeida/2003/04.25.17.12
6. Notes
Mark1
Empty Fieldsarchivingpolicy archivist callnumber contenttype copyholder copyright creatorhistory descriptionlevel dissemination documentstage doi edition group identifier isbn issn label lineage mirrorrepository nextedition nexthigherunit notes numberofvolumes orcid parameterlist parentrepositories previousedition previouslowerunit progress project readergroup readpermission resumeid rightsholder schedulinginformation secondarydate secondarykey secondarymark serieseditor session shorttitle sponsor subject tertiarymark url versiontype volume


Close