1. Kazumasa Aonashi* and Hisaki Eito Meteorological Research Institute, Tsukuba, Japan aonashi@mri-jma.go.jp July 27, 2011 IGARSS2011 Displaced Ensemble variational assimilation method to incorporate microwave imager TBs into a cloud-resolving model into a cloud-resolving model

2. Satellite Observation (TRMM) Infrared Imager SST, Winds Cloud Particles Frozen Precip. Snow Aggregates Melting Layer Rain Drops Radiation from Rain Scattering by Frozen Particles Radar Back scattering from Precip. Scattering Radiation 0℃0℃ Microwave Imager Cloud Top Temp. 10μm 3 mm-3cm （１００－１０ GH ｚ） 2cm １９ GH ｚ ８５ GHz

3. Cloud-Resolving Model used JMANHM （ Saito et al,2001) Resolution: 5 km Grids: 400 x 400 x 38 Time interval: 15 s Explicitly forecasts 6 species of water substances

4. Goal: Data assimilation of MWI TBs into CRMs Hydrological Model Cloud Reslv. Model + Data Assim System MWI TBs (PR) Precip.

5. OUTLINE IntroductionMethodology Ensemble-based Variational Assimilation (EnVA) Displacement error correction (DEC) Displacement error correction (DEC) Application results Case （ 2004/6/9) Typhoon CONSON(0404) Assimilation Results Impact on precipitation forecasts Summary & future directions

6. TMI 040609.OP37437

7. OUTLINE IntroductionMethodology Ensemble-based Variational Assimilation (EnVA) Displacement error correction (DEC) Displacement error correction (DEC) Application results Case （ 2004/6/9) Typhoon200404 Assimilation Results Impact on precipitation forecasts Summary & future directions

8. Methodology Ensemble-based Variational Assimilation (EnVA) (Lorenc 2003, Zupanski 2005) Displacement error correction (DEC) Data assimilation schemes

9. Why Ensemble-based method?: 200km 10km Heavy Rain AreaRain-free Area To estimate the flow-dependency of the error covariance Ensemble forecast error corr. of PT (04/6/9/22 UTC)

10. Why Variational Method ? Why Variational Method ? MWI TBs are non-linear function of various CRM variables. TB becomes saturated as optical thickness increases: TB depression mainly due to frozen precipitation becomes dominant after saturation. To address the non- linearity of TBs

11. Presupposition of Ensemble-based assimilation Obs. Analysis ensemble mean T=t0T=t1T=t2 Analysis w/ errorsFCST ensemble mean Ensemble forecasts have enough spread to include (Obs. – Ens. Mean)

12. Displacement error betw. Observation & Ensemble forecast Large scale displacement errors of rainy areas between the MWI observation and Ensemble forecasts Presupposition of Ensemble assimilation is not satisfied in observed rain areas without forecasted rain. AMSRE TB19v (2003/1/27/04z) Mean of Ensemble Forecast (2003/1/26/21 UTC FT=7h ）

13. Ensemble-based assimilation for observed rain areas without forecasted rain Ensemble-based assimilation for observed rain areas without forecasted rain Obs. Analysis ensemble mean T=t0T=t1T=t2 Analysis w/ errorsFCST ensemble mean Assimilation can give erroneous analysis when the presupposition is not satisfied. Signals from rain can be misinterpreted as those from other variables Displacement error correction is needed!

14. Displaced Ensemble variational assimilation method In addition to, we introduced to assimilation. The optimal analysis value maximizes ： Assimilation results in the following 2 steps: 1) DEC scheme to derive from 2)EnVA scheme using the DEC Ensembles to derive from

15. Fig. 1: CRM Ensemble Forecasts Displacement Error Correction Ensemble-based Variational Assimilation MWI TBs Assimilation method

16. DEC scheme: min. cost function for d Bayes’ Theorem can be expressed as the cond. Prob. of Y given ： We assume Gaussian dist. of ： where is the empirically determined scale of the displacement error. We derived the large-scale pattern of by minimizing (Hoffman and Grassotti,1996) ：

17. Detection of the large-scale pattern of optimum displacement We derived the large-scale pattern of from, following Hoffman and Grassotti (1996) ： We transformed into the control variable in wave space, using the double Fourier expansion. We used the quasi-Newton scheme (Press et al. 1996) to minimize the cost function in wave space. we transformed the optimum into the large-scale pattern of by the double Fourier inversion.

18. Fig. 1: CRM Ensemble Forecasts Displacement Error Correction Ensemble-based Variational Assimilation MWI TBs Assimilation method

19. EnVA: min. cost function in the Ensemble forecast error subspace Minimize the cost function Assume the analysis error belongs to the Ensemble forecast error subspace （ Lorenc, 2003): Forecast error covariance is determined by localization Cost function in the Ensemble forecast error subspace ：

20. Calculation of the optimum analysis Detection of the optimum by minimizing Transform of using eigenvectors of S ： Minimize the diagonalized cost function Approximate the gradient of the observation with the finite differences about the forecast error: Following Zupanski (2005), we calculated the analysis of each Ensemble members, from the Ensemble analysis error covariance.

21. Application results Case （ 2004/6/9) Typhoon CONSON (0404) Assimilation Results Impact on precipitation forecasts

22. Case （ 2004/6/9/22 UTC) TY CONSON 1) Assimilate TMI TBs 1) Assimilate TMI TBs (10v, 19v, 21v) (10v, 19v, 21v ) 2) 100 member Ensemble (init. 04/6/9/15 UTC ： FG) TMI TB19 ｖ RAM (mm/hr)

23. TB19v from TMI and CRM outputs FG ： First guess DE ： After DEC ND: NoDE+ EnVA CN ： DE+ EnVA TMI

24. RAM and Rain mix. ratio analysis (z=930m) FGDE RAM NDCN

25. RH(contours) and W(shades) along N-S FGDE CNND Ｓ Ｎ Ｓ Ｎ ＳＮ M

26. CRM Variables vs. TBc at Point M FG DE Qr(930m) vs.TB19v RTW(3880m) vs.TB21v vs.TB21v

27. Hourly Precip. forecasts (FT=0-1 h) 22-23Z 9th RAM DE CN ND FG

28. Hourly Precip. Forecasts (FT=3-4 h) 01-02Z 10th RAM DE CN ND FG

29. Summary Ensemble-based data assimilation can give erroneous analysis, particularly for observed rain areas without forecasted rain. In order to solve this problem, we developed the Ensemble-based assimilation method that uses Ensemble forecast error covariance with displacement error correction. This method consisted of a displacement error correction scheme and an Ensemble-based variational assimilation scheme.

30. Summary We applied this method to assimilate TMI TBs (10, 19, and 21 GHz with vertical polarization) for a Typhoon case (9th June 2004). The results showed that the assimilation of TMI TBs alleviated the large-scale displacement errors and improved precip forecasts. The DEC scheme also avoided misinterpretation of TB increments due to precip displacements as those from other variables.

31. Forecast error corr. of W (04/6/9/15z 7h fcst) Heavy rain (170,195) Weak rain (260,210) Rain-free (220,150) 200 km Severe sampling error for precip- related variables

32. Thank you for your attention. End

33. Ensemble-based Variational Assimilation Method Why Ensemble-based Assimilation method?: method?: To address the flow-dependency of the error covariance Why Variational Assimilation Method ? ： To address the non-linearity of TBs

34. Why Ensemble-based method?: Ensemble forecast corr. of PT (04/6/9/22 UTC) 200km 10km 1000 km Heavy Rain AreaRain-free Area To address the flow-dependency of the error covariance

35. Presupposition of Ensemble-based assimilation Ensemble forecasts have enough spread to include (Obs. – Ens. Mean) Assimilation can give erroneous analysis when the presupposition is not satisfied.

36. Cloud-Resolving Model used JMANHM （ Saito et al,2001) Resolution: 5 km Grids: 400 x 400 x 38 Time interval: 15 s Initial and boundary data JMA’s operational regional model Basic equations : Hydrostatic primitive Precipitation scheme: Moist convective adjustment + Arakawa-Schubert + Large scale condensation Resolution: 20 km Grids: 257 x 217 x 36

37. Explicit cloud microphysics scheme based on bulk method （ Lin et al.,1983; Murakami, 1990; Ikawa and Saito, 1991 ） The water substances are categorized into 6 water species (water vapor, cloud water, rain, cloud ice, snow and graupel) Explicitly predicting the mixing ratios and the number concentrations of frozen particles Cloud Microphysical Scheme

38. Why EnVA? Emission & Scattering signals vs. hydrometers Convective rain (Jan. 27, 2003) Emission Singals At 18 GHz τ ∝ LN(Ts-TB) Ts-TB= Ts(1-εs)exp(-2τ) Scattering Singals At 89 GHz τ ∝ LN(TB/TBflh) TB=TBflh Exp(-τ)

39. Fig. 1: CRM Ensemble Forecasts Displacement Error Correction Ensemble-based Variational Assimilation MWI TBs Assimilation method

40. Post-fit residuals FG ： DE ： ND: CN ： LN: DE+ EnVA. 1st Jx=24316.6 Jb=0 Jo=24316.6 Jx= 9435.2 Jb=0 Jo= 9435.2 Jx= 4105.4 Jb= 834.5 Jo= 3270.9 Jx= 2431.9 Jb= 290.9 Jo= 2141.0 Jx= 6883.0 Jb= 14.5 Jo= 6868.4