Precipitation Remote Sensing Huade Guan Prepared for Remote Sensing class Earth & Environmental Science University of Texas at San Antonio November 14, 2005
Outline Background Remote sensing technique for estimating precipitation, and related sensors NEXRAD Testing and improving NEXRAD products Future mission
Precipitation physics http://www.uwsp.edu http://eesc.columbia.edu/courses/ees/slides/climate/
www.gc.maricopa.edu Precipitation physics http://www.synthstuff.com/mt/archives/flickr-lenticular-cloud.jpg
Precipitation processes http://www.uwsp.edu http://rsd.gsfc.nasa.gov/rsd/images/georges/georgesms_md.jpg cold front http://www.mvinstitute.org warm front Occluded front http://www.uwsp.edu/geo/faculty/lemke/geog101/lecture_outlines/08_precipitation_processes.html
Gauge measurement http://www.hubbardbrook.org/yale/watersheds/w6/rain-gauge-stop/precipitation.htm http://www.usatoday.com/weather/wtipgage.htm Problems of gauge measurement: 1) Limited spatial coverage 2)
Four types of mapping approaches (examples) Information incorporated Spatial covariance No Yes Physical process No Yes Theissen polygon, & inverse square distance Regression, e.g., P-Z Kriging Cokriging (P-Z) & De-trended residual kriging
Precipitation remote sensing Satellite-based Geostationary (e.g., GOES) Polar orbiting (e.g., AVHRR, TRMM) Ground-based NEXRAD
VIS/IR technique Outgoing Longwave Radiation Basis: Precipitation leading to outgoing longwave radiation different from normal background Empirical relationship: P~OLR Example: IR bands of AVHRR or NOAA-series satellites for OLR, explained 40% of the areally average rainfall variability.
VIS/IR technique GOES Precipitation Index (GPI) Basis: cold cloud-top temperature leads to precipitation For pixels of cloud-top temperature (CCTs) less than 235 K are classified as raining pixel, and assigned a rainfall rate of 3 mm/hr Reproduce climate-scale precipitation patterns for tropics and sub-tropics But problematic for orographic and high-latitude precipitation
VIS/IR technique Bristol Algorithms (e.g., PERMIT: Polar-Orbiter Effective Rainfall Monitoring Integrative Technique) rain days based on the threshold IR brightness temperature Spatially variable mean-rain-per-day (from other sources) RAINSAT Use both visible and near-infrared Trained the model using radar observations
PERSIANN Products based on GOES infrared brightness temperature http://hydis8.eng.uci.edu/persiann/
Passive microwave technique Basis: precipitation-size ice particles and raindrops can scatter microwave and reduce the bulk emissivity of the cloud. 85.5 GHz brightness temperature SSM/I algorithms Empirical relationship
RADAR technique http://www.everythingweather.com/weather-radar/principles.shtml
TRMM RADAR TRMM PR sensor uses radar frequencies of 13.796 and 13.802 GHz horizontal resolution = 4.3 km at nadir measurements sensitivity better than 0.5 mm/h measures rain from the ground to an altitude of 15 km a vertical ("range") resolution of 250 m. provides 3-dimensional rainfall distribution
Next t Generation Weather Radar ar WSR-88D (NEXRAD( NEXRAD) Standard Z = 300 R 1.4 Tropical Z = 250 R 1.2 Unit! http://www.everythingweather.com/weather-radar/principles.shtml
First deployed: in 1988 Wavelength: 10cm Spatial Resolution (km): ~ 4 Temporal Resolution: 6-10 minutes 160 Radars
Virga effect, range degradation and beam blockage The radar will complete one volume scan (nine elevation scans) every six minutes. The radar will complete one volume scan (14 elevation scans) every five minutes
Bright band contamination http://apollo.lsc.vsc.edu/classes/remote/lecture _notes/radar/conventional/bright_band.html http://grappa.meteo.mcgill.ca/bright_band.html
NEXRAD rainfall products (4 km and hourly) Stage I - Hourly digital precipitation (HDP) Stage II - HDP merge with gauges Stage III - Mosaicked Stage II cover a RFC area or MPE (Multi-sensor Precipitation Estimator) Stage IV Stage IV - Mosaicked Stage III / MPE for continental U.S.
(Richard Fulton, Dong-Jun Seo, Jay Breidenbach, 2002)
Stage III/MPE in 13 RFCs http://dipper.nws.noaa.gov/hdsb/data/nexrad/wgrfc_stageiii.html
DATABASE and Visualization ArcIMS HTML viewer and JAVA viewer Data can be downloaded: ftp://epscor-data@ftp.ees.nmt.edu/
NEXRAD rainfall products testing
A physically based parsimonious approach (ASOADeK) for NEXRAD rainfall downscaling 4km 1km
Physical process (1) Orographic effects on precip. Orographic lifting, & hindrance Reduction in virga effect We use cos (α-ω) to approximate terrain aspect effects wind direction:ω wind T terrain aspect: α Elevation (Z) T P (windward) > P( leeward) P (low Z) < P( high Z) P (low Z) < P( high Z) terrain aspect
Physical process (2) Atmospheric effects on precipitation How does this heterogeneous atmospheric moisture distribution (or gradient in atmospheric moisture) influence precipitation? May Precip. Map We use geographic coordinates (Longitude or X, and Latitude or Y) to capture the effect of gradient in atmospheric moisture on precipitation GOES East 4-km, infrared imagery 2001.05.04 Study area
P Auto-search orographic and Regression: atmospheric effects aspect = b + b X + b Y +... + b Z + b cos( α ω) 0 1 2 3 4 moist. flux dir. gradient in moist., elevation, aspect & moist. flux direct. Data: Gauge precip: X, Y, P; Elev. DEM: X, Y, Z, α ; But what about moisture flux direction, ω? cos( α ω) Let : b b 4 4 cosω = b sinω = b = cosα cosω + sinα sinω 5 6
Future GPM's two instruments: Dual-frequency Precipitation Radar (DPR), and the GPM Microwave Imager (GMI)
Piloted UAVs GPM Primary Satellite Radar/Radiometer Prototype Instruments Ground validation Meteorology-Microphysics Aircraft Canada England Germany NASA Land Spain Italy South Korea NASA KSC France (Niger-Benin) India Japan Taiwan NASA Ocean Brazil Supersite Australia Regional Raingauge Site Both Supersite & Raingauge Site