1 / 17

Validation of the ENVISAT RA2 rain flag Jean Tournadre, Laboratoire d ’Océanographie Spatiale

Validation of the ENVISAT RA2 rain flag Jean Tournadre, Laboratoire d ’Océanographie Spatiale. Main effect of Rain on e.m. signals: attenuation (frequency dependant) Attenuation one order of magnitude larger at Ku band than at S band (S band unaffected)

loman
Download Presentation

Validation of the ENVISAT RA2 rain flag Jean Tournadre, Laboratoire d ’Océanographie Spatiale

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Validation of the ENVISAT RA2 rain flagJean Tournadre, Laboratoire d ’Océanographie Spatiale • Main effect of Rain on e.m. signals: attenuation (frequency dependant) • Attenuation one order of magnitude larger at Ku band than at S band (S band unaffected) • RAIN FLAG: detection of occurrences for whichs0ku is significantly attenuated compared to s0S where f is the “rain free” Ku/S band relation and rms the rms around this relation To avoid false alarm: Liquid water content >0.25 mm

  2. Validation of rain flag • Statistical analysis of s0 Ku and S data and their corrections for cycles 10 11 and 12 (October to December 2002) • Estimation of the Ku/S band relation analysis of its variability • Comparison with theoretical relation • Rain flagged samples

  3. Data screening The f relation = “rain free” relation between Ku and S s0 measurements To avoid any problems with heavy rain, sea ice, land contamination, instrumental problems the following criteria have been used to screen the data flags: • altim_landocean_flag, radio_landocean_flag, mwr_instr_flags, • instr_flags , ocean_retrk_qua_flags, s_ocean_retrk_qua_flags set to 0. Geophysical values: • s0Ku>0 s0S>0 • mwr_liq_vapour_cont<50.10-2kg/m² off_nad_ang_wvform<1deg² • ku_peak <0.19 s_peak <0.19 • 60°S<latitude<60°N

  4. Statistical analysis of s0 data s0 in IGDR defined by: November 2002 Mean s0Ku~ 11 dB rms ~1.7 dB Mean s0 S~10.5 dB rms~1.91 dB Different from the first IGDR (~2dB) The histograms comparable to those of JASON Very good stability in time December 2002

  5. s0 corrections (instrumental and atmospheric) Nov 2002 • Atmospheric correction includes cloud liquid water (from MWR) • Should not be used for rain flagging • Mean correction for KU ~0.13dB • negligible for S Dec 2002 Instrumental correction Mean correction for Ku band 2.5 dB and 3.13 dB for S band Almost constant (rms~0.05dB) Distribution skewed for Ku band

  6. Spatial variabilityof s0 Distribution of sample to sample Ds0 (Dx= 5.8km) for Ku and S band. Estimation of natural s0 variability Similar Ku and S band distributions (gaussian, 0 mean 0.15dB rms) Rms similar to Topex and Jason, but the distributions are less peaky

  7. Ku/C band relation : s0 bidimensional pdf Nov 2002 The data are screened for land, ice (using peakiness) and instrumental problems (~106 samples used). Between 7 and 13 dB good agreement with theoretical relation (with 5 dB offset for S band and 1dB for Ku band) No significant difference between the two month Dec 2002

  8. Ku/C band relation and rms Nov 2002 Dec 2002 • mean Ku/S band f relation and its rms computed from s0 data set. • f relation used to reject data if Ds0 > 2 rms new f1 estimated • A priori rms from Topex (green line) • Good agreement with theoretical relation with DsC=5dB DsKu=1dB • No difference except for very low wind (s0 >17dB) where the f relation has a larger slope than the theoretical relation (results from wave impact) • Very good stability

  9. Nov 2002 Distributions of Ds0=s0Ku –f(s0S) and Ds0/2 rms Skewed for Ds0 (influence of rain) Gaussian distribution for the normalized departure Point flagged for Ds0 /2rms<-1 and Lz>250mm Dec 2002

  10. 2D distribution of s0 Ku and S band : very large dispersion for high s0 (low winds), much larger than that of TOPEX and JASON. This results from a stronger influence of significant wave heights on s0 The f relation has been computed as a function of Ku band SWH Large differences for s0>15dB, no influence on rms Explains the difference between theory and observation Theoretical relation computed with SWH=1m

  11. Rain probability The RA2 data are screened for the flags and the geophysical value but not for the MWR Lz. Scale from 0 to 10% of flagged samples. Still, some problems remain with sea ice flag and land flag f1 November 2002 December 2002 f

  12. New products Relation between attenuation and freezing level and Brightness temperature at 23.8GHz We used the Wilheit et al (1977) model to compute the brightness temperature at 23.8GHz as a function of Ku band attenuation and freezing level height For rainy samples the freezing level is then inferred from the 23.8GHz temperature

  13. Example of mean freezing level estimated from rainy samples for November 2002. Good agreement with climatology

  14. Mean rain rate for November 2002 Estimated by inverting the attenuation in terms of rain rate Where a and b are frequency dependant coefficients and H the freezing level

  15. Mean rain rate for November 2002 using the freezing level, H, determined from 23.8 GHz temperature as a correction Larger rain rates at higher latitude

  16. Conclusion/perspective • Distributions of s0 at S and Ku band very stable in time. Mean value corrected since the first IGDR (~2.2dB) • Atmospheric correction always removed; • Good sample to sample variability distribution (similar to Topex and Jason but less peaky distribution) • Ku/S s02D distributions similar to those of JASON and TOPEX and very stable in time • f relation very similar to theoretical one up to 15dB (influence of SWH) • Number of flagged samples ~2% • Mean rain probability fields similar to the Topex and Jason ones • New products (Freezing level, rain rate fields)

  17. Model of rain/altimeter interaction

More Related