Скачать презентацию SOME EXPERIENCES ON SATELLITE RAINFALL ESTIMATION OVER SOUTH Скачать презентацию SOME EXPERIENCES ON SATELLITE RAINFALL ESTIMATION OVER SOUTH

7b3df1dd3e91fdddeb3d77082bd8a528.ppt

  • Количество слайдов: 41

SOME EXPERIENCES ON SATELLITE RAINFALL ESTIMATION OVER SOUTH AMERICA: HYDROESTIMATOR TECHNIQUE Daniel Vila (dvila@essic. SOME EXPERIENCES ON SATELLITE RAINFALL ESTIMATION OVER SOUTH AMERICA: HYDROESTIMATOR TECHNIQUE Daniel Vila ([email protected] umd. edu) ESSIC/CICS – Computer & Space Science Building - Bldg #224 College Park, MD 20742

Outlines • Something about satellite rainfall estimation. • Brief algorithm description of the South Outlines • Something about satellite rainfall estimation. • Brief algorithm description of the South American version of NOAA/NESDIS “Hydro-Estimator” satellite rainfall estimation and correction methodologies. • Image availability and convection activity. • Verification methods and some preliminary results over Del Plata basin. • Conclusions.

Satellite Quantitative Precipitation Estimation (QPE) 1970 s ------ Manual Procedures 1970 s 1980 s Satellite Quantitative Precipitation Estimation (QPE) 1970 s ------ Manual Procedures 1970 s 1980 s - 1990 s ---- Interactive Methods / some use of 1980 s - 1990 s microwave 2000 ------ Single channel (10. 7 mm) automatic algorithm 2000 2005 ------ GOES multi-channel and microwave 2005 2008 ------ Global Precipitation Measurement (GPM) 2008 Mission and GOES (postponed 2012? ) 20? ? ------ GPM and Hyperspectral (GIFTS) and GOES 20? ? Microwave?

Satellite Quantitative Precipitation Estimation (QPE) Why Use Satellite QPE? n Superior spatial coverage Offshore Satellite Quantitative Precipitation Estimation (QPE) Why Use Satellite QPE? n Superior spatial coverage Offshore coverage (tropical systems) n No beam block problems n n Consistency Differences in calibration from radar to radar n Radar range effects n Beam overshoot (especially stratiform precip) n replacement, but a companion to radar v Not a

Satellite Quantitative Precipitation Estimation (QPE) GOES-Based QPE: Theory Basis • Assumes that cloud-top temperature Satellite Quantitative Precipitation Estimation (QPE) GOES-Based QPE: Theory Basis • Assumes that cloud-top temperature ~ cloud-top height cloud evolution cloud-top thickness rainfall rate Strengths • 24/7 coverage every 15 minutes throughout North America (this is not true in South America!!) • High spatial resolution (~4 km) Weaknesses • Relationship between cloud-top properties and rain rate often does not hold, esp. for non-convective precipitation. • Cold cirrus can be mistaken for cumulonimbus.

Satellite Quantitative Precipitation Estimation (QPE) Microwave QPE Algorithms Ø Special Sensor Microwave/Imager (SSM/I)—available since Satellite Quantitative Precipitation Estimation (QPE) Microwave QPE Algorithms Ø Special Sensor Microwave/Imager (SSM/I)—available since 1987 Ø Advanced Microwave Sounding Unit-A (AMSU-A)—available since 1999 Ø Advanced Microwave Sounding Unit-B (AMSU-B)—available since 2000 Ø Tropical Rainfall Measuring Mission (TRMM) --- available since 1998

Satellite Quantitative Precipitation Estimation (QPE) Microwave QPE Algorithms Theory Satellite Quantitative Precipitation Estimation (QPE) Microwave QPE Algorithms Theory

Satellite Quantitative Precipitation Estimation (QPE) Satellite Quantitative Precipitation Estimation (QPE)

Satellite Quantitative Precipitation Estimation (QPE) Microwave-IR Blended Algorithm • Relationship between 10. 7 - Satellite Quantitative Precipitation Estimation (QPE) Microwave-IR Blended Algorithm • Relationship between 10. 7 - mm Tb and rain rate calibrated using SSM/I rain rate estimates “Best of both worlds”—combine robustness of MW estimates with availability of GOES data • Calibration updated every few hours for a 5 x 5 -degree region • Uses all operational adjustments (moisture, orography, etc. ) • Developed by F. J. Turk of NRL

Program for the Evaluation of High Resolution Precipitation Products (PEHRPP) International Precipitation Working Group Program for the Evaluation of High Resolution Precipitation Products (PEHRPP) International Precipitation Working Group (Working Group of CGMS) found this an interesting and relevant problem, and agreed to sponsor and advise the effort (http: //www. isac. cnr. it/~ipwg/) Hypotheses: • HRPP errors can be characterized by comparing them to independent observations from rain gauges and radars. • Errors of and differences between HRPP are meaningful, in that they can be systematically related to precipitation characteristics and/or algorithm methodology. • Improved HRPP can be derived by combining products or methods based on the observed errors and differences. • HRPP spatial and temporal variability is realistic on scales appropriate for scientific studies (e. g. , hydrology). • Numerical weather prediction forecasts of precipitation can be used to improve HRPP in some locations and times (e. g. , high latitudes).

Program for the Evaluation of High Resolution Precipitation Products (PEHRPP) Program for the Evaluation of High Resolution Precipitation Products (PEHRPP)

Program for the Evaluation of High Resolution Precipitation Products (PEHRPP) http: //www. cpc. ncep. Program for the Evaluation of High Resolution Precipitation Products (PEHRPP) http: //www. cpc. ncep. noaa. gov/products/janowiak/us_web. shtml

Hydroestimator algorithm description • Rain rate estimation: fully automated method using an empirical power-law Hydroestimator algorithm description • Rain rate estimation: fully automated method using an empirical power-law function that generates rainfall rates (mm/h) based on GOES-8 channel 4 brightness temperature and precipitable water (ETA model). • Moisture correction: precipitable water (PW) (integrated over the layer from surface to 500 h. Pa) and relative humidity (RH) (mean value between surface and 500 h. Pa. , in percentage) obtained by ETA model are applied to decrease rainfall rates in dry environments and increases them in the moist ones. • Orographic correction: a Digital Elevation Model of South America at the GOES scale combined with low-level winds to produce an orographic correction to the satellite rainfall rate distribution

Hydroestimator algorithm description • Screening method: This technique assumes that raining pixels are colder Hydroestimator algorithm description • Screening method: This technique assumes that raining pixels are colder than the mean of the surrounding pixels. • Clusterization: Statistical parameters (average temperature and standard deviation) of each system is used to produce adjustments resulting in more realistic patterns of rain, especially for small-scale convection (CPTEC algoritm). RAIN RATE ESTIMATION • Standardized temperature is defined as:

Hydroestimator algorithm description • Tơ < -1. 5 Conventive core: defined essentially by the Hydroestimator algorithm description • Tơ < -1. 5 Conventive core: defined essentially by the empirical power-law function corrected by PW. • Tơ = 0 “Non-core” precipitation: whose maximum value cannot exceed 12 mmh-1 and must be less than the fifth part of the convective rainfall for a given pixel • – 1. 5 < Tơ < 0 • Tơ > 0 pp = 0

Hydroestimator algorithm description Brigthness temperature - rainfall rate relationship for various values of precipitable Hydroestimator algorithm description Brigthness temperature - rainfall rate relationship for various values of precipitable water (mm) for convective core and for “non-core” precipitation

Hydroestimator algorithm description Ororaphic correction: east-southest wind flow produces enhaced precipitation upstream and less Hydroestimator algorithm description Ororaphic correction: east-southest wind flow produces enhaced precipitation upstream and less rainfall downstream

Image availability for southern hemisphere sector from 13 January 2003 to 28 February. Green Image availability for southern hemisphere sector from 13 January 2003 to 28 February. Green boxes represent available images

Image availability Image availability

Image availability NOAA PLANS SHIFT IN GEOSTATIONARY SATELLITE ORBIT TO IMPROVE WEATHER FORECAST COVERAGE Image availability NOAA PLANS SHIFT IN GEOSTATIONARY SATELLITE ORBIT TO IMPROVE WEATHER FORECAST COVERAGE OVER SOUTH AMERICA Continent Will Benefit from Emerging Global Earth Observation Network Jan. 18, 2006 — In Buenos Aires on Tuesday, NOAA, the Comisión Nacional de Activades Espaciales, and the World Meteorological Organization announced news of the repositioning of GOES-10. Shifting the spacecraft from its current position above the equator in the West to a new spot in orbit will greatly improve environmental satellite coverage of the Western Hemisphere, especially over South America. The repositioning is planned for October 2006 pending the successful launch of GOES-N, NOAA's new geostationary satellite, and the continued operation of GOES-12. (extracted from NOAA News Online - http: //www. noaanews. noaa. gov/stories 2006/s 2561. htm

 Verification methods Verification methods

 Verification methods Verification methods

 Preliminary results EXPERIMENT DESIGN • Period: January - February 2003 • Region: Del Preliminary results EXPERIMENT DESIGN • Period: January - February 2003 • Region: Del Plata basin (40°S - 20 °S, 60°W - 35 °W) • Image availability: more than 28 images in 24 hours. 27 cases were selected • Ground truth: Raingauges (INMET meteorological stations, PCDs and other networks) • Integration method: cubic splines time interpolation.

 Preliminary results BIAS: mean value= -4. 02 mm Preliminary results BIAS: mean value= -4. 02 mm

 Preliminary results RMSE: mean value = 12. 7 mm Preliminary results RMSE: mean value = 12. 7 mm

 Preliminary results POD & FAR: POD mean value = 0. 83 FAR mean Preliminary results POD & FAR: POD mean value = 0. 83 FAR mean value = 0. 05

 Case study (1) 21 January 2003 Case study (1) 21 January 2003

 Case study (1) 21 January 2003 (CPTEC algorithm) NESDIS algorithm (available in INTERNET) Case study (1) 21 January 2003 (CPTEC algorithm) NESDIS algorithm (available in INTERNET) Overall statistics for 20 -21 January 2003. Underestimation is present in all intervals for both algorithms.

 Case study (1) 21 January 2003 Case study (1) 21 January 2003

 Case study (2) 10 -11 February 2003 Case study (2) 10 -11 February 2003

 Case study (2) 12 February 2003 (Diario El Pais - Uruguay) Case study (2) 12 February 2003 (Diario El Pais - Uruguay)

 Case study (2) 10 -11 February 2003 Case study (2) 10 -11 February 2003

Cuareim River Basin • 24 -hours rainfall estimation using HE technique (NESDIS version) • Cuareim River Basin • 24 -hours rainfall estimation using HE technique (NESDIS version) • Period: March 1 st 2002 – July 31 st 2003 • Calibration data: GTS available raingauges • Validation data: CTMSG raingauge network

Cuareim River Basin Scatterplot of Observed vs. Estimated mean areal rainfalll Cuareim River Basin Scatterplot of Observed vs. Estimated mean areal rainfalll

Cuareim River Basin • This correction algorithm takes into account the difference between rain Cuareim River Basin • This correction algorithm takes into account the difference between rain gauges and the HE estimation for a given rain gauge network Schematic procedure of the best adjusted value (MVE). Rainfall data is compared with a nine pixels kernel centered in the rain gauge location

Cuareim River Basin Scatterplot of Observed vs. Estimated mean areal rainfall corrected with rain Cuareim River Basin Scatterplot of Observed vs. Estimated mean areal rainfall corrected with rain gauges values.

Statistical Summary • The results of the corrected spatial precipitation estimates improve noticeably with Statistical Summary • The results of the corrected spatial precipitation estimates improve noticeably with respect to those without correction.

 Case study (3) Case study (3)

 Case study (3) Case study (3)

 Conclusions • The main purpose of this work is to present the recent Conclusions • The main purpose of this work is to present the recent improvements of the Auto-Estimator Algorithm (Hydroestimator) and the application of this technique in Del Plata basin (South America). • The main difference between the South American model and the one for North America is the image availability. Gaps up to three hours in South America imagery may be a very important factor in the accuracy of the estimations. • The errors involved in these kind of techniques were evaluated during January-February 2003 (summer) and the case study presented. • More detailed statistics (i. e. warm and cold tops discrimination and seasonal behavior) will be carried out in the future.

MUCHAS GRACIAS !!! THANK YOU !!! MUCHAS GRACIAS !!! THANK YOU !!!