Download - Application of Radar QPE - thamesrevival.ca
![Page 1: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/1.jpg)
Application of Radar QPE
Jack McKeeDecember 3, 2014
![Page 2: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/2.jpg)
TopicsTopics
• Context• Precipitation Estimation TechniquesPrecipitation Estimation Techniques• Study Methodology• Preliminary Results• Future Work• Future Work• Questions
![Page 3: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/3.jpg)
IntroductionIntroduction
• Accurate precipitation data is required for a number of different applications ppincluding:– Flood forecasting and warningFlood forecasting and warning– Drought monitoring – Meteorological now-casting– Meteorological now-casting– Post-storm analysis
Effect of climate change on precipitation patterns– Effect of climate change on precipitation patterns– Etc.
![Page 4: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/4.jpg)
HydrologyHydrology
• Hydrologists rely on precipitation fields to forecast runoffs
![Page 5: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/5.jpg)
Hydrological modelingHydrological modeling
• Hydrological model attempts to model the complex physical characteristics of a basin p p yin order to determine stream flow
“Rainfall is the main and vital [input] for h d l i l d li d i f ll t i thydrological modeling and rainfall uncertainty
dominates the uncertainty of the model” Golding (2009)
![Page 6: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/6.jpg)
GoalGoal
Develop a reliable automated process for obtaining precipitation estimates using state-obtaining precipitation estimates using state-of-the-art techniques for the acquisition and
ti ti f ti ll di t ib t d i f llestimation of spatially distributed rainfall fields
![Page 7: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/7.jpg)
TopicsTopics
• Context• Precipitation Estimation TechniquesPrecipitation Estimation Techniques• Study Methodology• Preliminary Results• Future Work• Future Work• Questions
![Page 8: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/8.jpg)
Precipitation EstimationPrecipitation EstimationW b St & R i G E t R d St tiWaubuno Stream & Rain Gauge
(UTRCA)Exeter Radar Station
(Environment Canada)
![Page 9: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/9.jpg)
Rain Gauges DefinitionRain Gauges - Definition
• Most widely used technique for precipitation estimationp p
R i th d th f i f ll• Rain gauges measure the depth of rainfall over a set time for a given location
P id t i t t• Provide accurate point measurements
![Page 10: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/10.jpg)
Rain Gauges ErrorsRain Gauges - Errors
• Two Critical Sources of Error:– inability of point measurements to accurately
characterize the spatial distribution of the rainfall field– systematic and calibration errors
• Significantly impact the ability to use rain• Significantly impact the ability to use rain gauge measurement to develop an
d di f h di ib i funderstanding of the distribution of precipitationp p
![Page 11: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/11.jpg)
Rain Gauges ErrorsRain Gauges - Errors
![Page 12: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/12.jpg)
Rain Gauges ErrorsRain Gauges - Errors
![Page 13: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/13.jpg)
Rain Gauges ErrorsRain Gauges - Errors
![Page 14: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/14.jpg)
Rain Gauges ErrorsRain Gauges - Errors
![Page 15: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/15.jpg)
Radar DefinitionRadar - Definition
• Transmits pulses of microwave to detect precipitation particles in the atmospherep p p p
![Page 16: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/16.jpg)
Radar DefinitionRadar - Definition
• Actually measures reflectivity not rainfall• Indirectly measures rainfall intensityIndirectly measures rainfall intensity
![Page 17: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/17.jpg)
Radar ErrorsRadar - Errors
• Radar suffers from numerous sources of error which influence the accuracy of the ydepth measurements
• Creutin et al (2000) characterized three• Creutin et al. (2000) characterized three major sources of radar error for QPE: • Electronic instability and mis-calibration of the radar
systemB• Beam geometry
• Fluctuation in atmospheric conditions
![Page 18: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/18.jpg)
Radar ErrorsRadar - Errors
![Page 19: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/19.jpg)
Radar ErrorsRadar - Errors
• Radome wetting caused underestimation of rainfallof rainfall accumulation of 100 mm100 mm (Boodoo et al. 2014)
![Page 20: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/20.jpg)
ComparisonComparison
Ki t l 2008 difi d f R bbi t l 2004Kim et al. 2008 modified from Robbins et al. 2004
![Page 21: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/21.jpg)
Merging MethodsMerging Methods
• Developed to take advantage of each individual instruments strength while gminimizing their respective weaknesses
“The combined use of radar and rain gauges to measure rainfall is superior to the use of
either separately”either separately
Wil (1970 495)Wilson (1970, p. 495)
![Page 22: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/22.jpg)
AssumptionsAssumptions
1. Gauge measurements are accurate for the gauges location
2. Radar successfully measures relative spatial and temporal variability's of precipitationy
3. Gauge and radar measurements are valid for the same location in time and spacethe same location in time and space
4. Relationships based on comparisons between gauges and radar(s) are valid for othergauges and radar(s) are valid for other locations in space and time
![Page 23: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/23.jpg)
TopicsTopics
• Context• Precipitation Estimation TechniquesPrecipitation Estimation Techniques• Study Methodology• Preliminary Results• Future Work• Future Work• Questions
![Page 24: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/24.jpg)
ObjectivesObjectives
• Study broken down into two parts:1. Compare the accuracy of gauge-radar p y g g
merging schemes in the UTRb;2 Assess the impact on accuracy of predicted2. Assess the impact on accuracy of predicted
flows in the Thames River and its tributaries
• Implement the “best” estimation ptechnique in a near-real time automated processprocess
![Page 25: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/25.jpg)
ObjectivesObjectives
• Study broken down into two parts:1. Compare the accuracy of gauge-radar p y g g
merging schemes in the UTRb;2 Assess the impact on accuracy of predicted2. Assess the impact on accuracy of predicted
flows in the Thames River and its tributaries
![Page 26: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/26.jpg)
Study MethodologyStudy Methodology
• Rainfall estimations are compared against an independent verification network passumed to represent the “true” rainfall fieldfield
• Previous study has shown several factors fto have an influence on the overall
accuracy including:y g– Rain gauge density– Temporal resolution of correctionp– Storm type
![Page 27: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/27.jpg)
MFBMFB
• Mean Field Bias (MFB) correction• Changes the multiplicative a factor in theChanges the multiplicative a factor in the
Z-R relationshipC ti i li d l t ti• Correction is applied evenly to entire spatial domain
N
N
iiG
C 1bRZ
N
iiR
C
1
baRZ i 1
![Page 28: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/28.jpg)
BSABSA
• Brandes Spatial Adjustment (BSA)• Assumes radar bias is spatially dependentAssumes radar bias is spatially dependent
due to atmospheric and conditionsU B bj ti l i (B• Uses Barnes objective analysis (Barnes, 1964) to distributed correction factors
dWT
2
N
iii GWT
F 1*
EP
WTi exp
N
ii
i
WTF
1
11
![Page 29: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/29.jpg)
LBLB
• Local Bias (LB) correction with ordinary krigingg g
• Adopts the geostatistical method of kriging to distributed correction factorsto distributed correction factors
• The regionalized variable is the correction factor at the gauge location which describes radar bias at discrete locationsdescribes radar bias at discrete locations across the radar field
![Page 30: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/30.jpg)
CMCM
• Conditional Merging (CM) also known as kriging with radar based errorg g
• Assumes radar observation produces a true field of unknown values while the raintrue field of unknown values, while the rain gauges produce an unknown field of true values
.iKiiR sRsRs .P. iRiKi ssGsrecipCorr
![Page 31: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/31.jpg)
Study AreaStudy Area
• Upper Thames River Basin (UTRb)( )
• 3421 km2
D li t d i t 34• Delineated into 34 sub-basins
![Page 32: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/32.jpg)
Data Description Rain GaugesData Description – Rain Gauges
• Upper Thames River Conservation Authority operates 20 tipping bucket rain y p pp ggauges
• Record on an hourly time interval• Record on an hourly time interval• 14 gauges selected for correction • 4 gauges selected as an independent
verification networkverification network• 2 were located in the shadow zone
![Page 33: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/33.jpg)
Data Description Rain GaugesData Description – Rain Gauges
![Page 34: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/34.jpg)
Data Description RadarData Description - Radar
• Radar provided by Environment Canada’s meteorological research centerg
• Corrected QPE is only provided up to a range of 120 kmrange of 120 km
![Page 35: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/35.jpg)
Data Description RadarData Description - Radar
![Page 36: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/36.jpg)
TopicsTopics
• Context• Precipitation Estimation TechniquesPrecipitation Estimation Techniques• Study Methodology• Preliminary Results• Future Work• Future Work• Questions
![Page 37: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/37.jpg)
VariabilityVariability
Hourly accumulation for each estimation technique September 10, 2014, 22:00 (UTC)p , , ( )
![Page 38: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/38.jpg)
RMSERMSE
![Page 39: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/39.jpg)
MREMRE
![Page 40: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/40.jpg)
Gauge SensitivityGauge Sensitivity
![Page 41: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/41.jpg)
Temporal SensitivityTemporal Sensitivity
![Page 42: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/42.jpg)
TopicsTopics
• Context• Precipitation Estimation TechniquesPrecipitation Estimation Techniques• Study Methodology• Preliminary Results• Future Work• Future Work• Questions
![Page 43: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/43.jpg)
Future WorkFuture Work
• Test the effect of storm type (based on seasonal differences))
• Assess the impact of the merging methods on simulated flows in the UTRb using aon simulated flows in the UTRb using a semi-distributed hydrological model
![Page 44: Application of Radar QPE - thamesrevival.ca](https://reader031.vdocuments.us/reader031/viewer/2022020621/61eae9d52d166b59d14cd40e/html5/thumbnails/44.jpg)
AcknowledgementsAcknowledgements