subroutine nlte_ch4(ngrid,nlayer,nq,pplay,pplev,pt,vmrch4,eps23,eps33,eps76) use comgeomfi_h implicit none !================================================================== ! Purpose ! ------- ! Calculation of epsilon - non LTE efficiency factor ! For CH4 heating and cooling rate calculation ! Cf Strobel et al 1996 ! ! Inputs ! ------ ! ngrid Number of vertical columns ! nlayer Number of layers ! pplay(ngrid,nlayer) Pressure layers ! pt ! ! Outputs ! ------- ! ! eps_sw ! efficiency factor for heating rate ! eps_lw ! efficiency factor for cooling rate ! ! Authors ! ------- ! Tanguy Bertrand (2015,2022) ! !================================================================== #include "dimensions.h" #include "dimphys.h" #include "comcstfi.h" #include "surfdat.h" #include "comvert.h" #include "callkeys.h" #include "tracer.h" !----------------------------------------------------------------------- ! Arguments INTEGER ngrid, nlayer, nq REAL pplay(ngrid,nlayer) REAL pplev(ngrid,nlayer) REAL pt(ngrid,nlayer) REAL vmrch4(ngrid,nlayer) !----------------------------------------------------------------------- ! Local variables INTEGER l,ig REAL cbol ! Boltzman constant REAL avog ! avogadro constant REAL beta23 REAL beta33 REAL nd(ngrid,nlayer) ! density number cm-3 REAL na(ngrid,nlayer) ! column density number cm-2 REAL rho(ngrid,nlayer) ! density of atmosphere REAL kn3,k3,k2n4,k3n4,k4 REAL An3,A3,A2n4,A3n4,A4 REAL dphin3,dphi3,dphi2n4,dphi3n4,dphi4 REAL phin3(ngrid,nlayer) REAL phi3(ngrid,nlayer) REAL phi2n4(ngrid,nlayer) REAL phi3n4(ngrid,nlayer) REAL phi4(ngrid,nlayer) REAL phi4T(ngrid,nlayer) ! DATA from Strobel et al 1996 REAL, DIMENSION(22) :: gam2 ! band escape fonction 2.3 REAL, DIMENSION(22) :: Naval2 ! column density number 2.3 REAL, DIMENSION(22) :: gam3 ! band escape fonction 3.3 REAL, DIMENSION(22) :: Naval3 ! column density number 3.3 REAL gammafi23(nlayer),gammafi33(nlayer) ! Band espace fonction REAL gammafi23mod(nlayer),gammafi33mod(nlayer) ! Band espace fonction REAL vmr REAL k4zT ! function to calculate k4 depending on T and vmr REAL log_square ! log function REAL log_affine ! log function REAL ratio_affine ! linear approx function to calculate ratio for gamma calculation REAL ratio_square ! REAL ratio_sq,ratio_af ! Output REAL,intent(out) :: eps23(ngrid,nlayer) REAL,intent(out) :: eps33(ngrid,nlayer) REAL,intent(out) :: eps76(ngrid,nlayer) ! Option simplified coefficient REAL klw, ksw, eps_sw1Pa !----------------------------------------------------------------------- eps23(:,:)=1. eps33(:,:)=1. eps76(:,:)=1. IF (strobel) then ! Constantes cbol=1.3806488e-23 !Boltzman constant avog=6.022141e23 ! band 3 k3=2.8e-11 !cm-3 A3=25.2 dphi3=k3/A3 ! band n3 kn3=2.8e-11 !cm-3 An3=27.3 dphin3=kn3/An3 ! band 3n4 k3n4=1.16e-11 !cm-3 A3n4=2.12 dphi3n4=k3n4/(3.*A3n4) ! band 4 !! Here we use function k4zT instead of constant k4 k4=4.5e-16 !cm-3 A4=2.12 dphi4=k4/A4 ! band 2n4 k2n4=1.16e-11 A2n4=2.12 dphi2n4=k2n4/(2.*A2n4) ! 2.3 band : beta23=0.09 ! 3.3 band : beta33=0.13 ! Escape Function Gamma (Data Strobel et al) : ! 100 K 2.3 Naval2=(/1.00000000e+10,1.30000000e+15,2.92766055e+15,6.59322793e+15,1.48482564e+16,3.34389650e+16, & 7.53061067e+16,1.69592860e+17,3.81931020e+17,8.60126447e+17,1.93704482e+18,4.36231516e+18, & 9.82413694e+18,2.21244140e+19,4.98252108e+19,1.12208695e+20,2.52699209e+20,5.69090388e+20, & 1.28161806e+21,2.88626357e+21,6.50000000e+21,1.00000000e+24/) gam2=(/1.00000000e+00,1.00000000e+00,1.00000000e+00,1.00000000e+00,1.00000000e+00,9.90000000e-01, & 9.80000000e-01,9.50000000e-0,9.00000000e-01,8.00000000e-01,7.00000000e-01,5.00000000e-01, & 3.50000000e-01,2.20000000e-01,1.20000000e-01,6.50000000e-02,3.20000000e-02,1.40000000e-02, & 6.20000000e-03,2.50000000e-03,1.05000000e-03,4.83227951e-06/) ! 100 K 3.3 Naval3=(/1.00000000e+10,1.30000000e+15,2.92766055e+15,6.59322793e+15,1.48482564e+16,3.34389650e+16, & 7.53061067e+16,1.69592860e+17,3.81931020e+17,8.60126447e+17,1.93704482e+18,4.36231516e+18, & 9.82413694e+18,2.21244140e+19,4.98252108e+19,1.12208695e+20,2.52699209e+20,5.69090388e+20, & 1.28161806e+21,2.88626357e+21,6.50000000e+21,1.00000000e+24/) gam3=(/1.00000000e+00,9.80000000e-01,9.00000000e-01,8.50000000e-01,7.00000000e-01,5.00000000e-01, & 3.00000000e-01,1.50000000e-01,1.00000000e-01,6.00000000e-02,4.00000000e-02,2.30000000e-02, & 1.45000000e-02,9.00000000e-03,5.00000000e-03,3.00000000e-03,1.50000000e-03,8.00000000e-04, & 3.80000000e-04,1.80000000e-04,8.30000000e-05,6.81737544e-07/) DO ig=1,ngrid phi3n4(ig,:)=1. phi2n4(ig,:)=1. phin3(ig,:)=1. phi4(ig,:)=1. nd(ig,:)=1. na(ig,:)=1. !write(*,*) 'TB22 pplev=',pplev(1,:) !write(*,*) 'TB22 tlev=',pt(1,:) DO l=1,nlayer !!calculation of density rho(ig,l)=pplay(ig,l)/(r*pt(ig,l)) !!calculation of column number density for ch4 (cm-2) na(ig,l)=vmrch4(ig,l)/100./(mmol(igcm_n2)*1.e-3)* & pplev(ig,l)/g*avog*1.e-4 !!calculation of atmospheric number density (cm-3) for each layer nd(ig,l)=rho(ig,l)/(mmol(igcm_n2)*1.e-3)*avog*1.e-6 !!calculation of the phi : phin3(ig,l)=dphin3*nd(ig,l) phi3(ig,l)=dphi3*nd(ig,l) phi2n4(ig,l)=dphi2n4*nd(ig,l) phi3n4(ig,l)=dphi3n4*nd(ig,l) phi4(ig,l)=dphi4*nd(ig,l) !! Getting phi4 depending on T and vmr vmr=vmrch4(ig,l)/100. phi4T(ig,l)=k4zT(pt(ig,l),vmr)/A4*nd(ig,l) ENDDO !write(*,*) 'TB22 nd=',nd(1,:) !write(*,*) 'TB22 vmr=',vmrch4 !write(*,*) 'TB22 phi4T=',phi4T !! interpolation of na values in cm-2 from Strobel values for !the 100K model: CALL interp_line(Naval2,gam2,22,na(ig,:),gammafi23,nlayer) CALL interp_line(Naval3,gam3,22,na(ig,:),gammafi33,nlayer) !write(*,*) 'TB22rad na = ',na(1,:) !write(*,*) 'TB22rad gammafi = ',gammafi23 DO l=1,nlayer !! Getting ratio depending on temperature and n (cm-2) ratio_sq=ratio_square(pt(ig,l),na(ig,l)) ratio_af=ratio_affine(pt(ig,l),na(ig,l)) !! Modified gamma depending on T gammafi23mod(l)=min(ratio_sq*gammafi23(l),1.) gammafi33mod(l)=min(ratio_af*gammafi33(l),1.) !! Heating : cf Strobel 1996 and Zalucha 2011 ! eps 2.3 eps23(ig,l)=phin3(ig,l)/(phin3(ig,l)+gammafi23mod(l))* & (beta23+phi3n4(ig,l)/(phi3n4(ig,l)+1)* & phi4T(ig,l)/(phi4T(ig,l)+gammafi23mod(l))* & (3.*phi2n4(ig,l)+1.)/(1.+phi2n4(ig,l))*(1-beta23)/3.) ! eps 3.3 eps33(ig,l)=phi3(ig,l)/(phi3(ig,l)+gammafi33mod(l))* & (beta33+phi2n4(ig,l)/(phi2n4(ig,l)+1)* & phi4T(ig,l)/(phi4T(ig,l)+gammafi33mod(l))* & (1-beta33)) ! Cooling from Zalucha et al 2013 eq10 eps76(ig,l)=1./(1.+gammafi23mod(l)/(2.*phi4T(ig,l))) ENDDO !write(*,*) 'TB22rad eps23 = ',eps23(1,:) !write(*,*) 'TB22rad eps33 = ',eps33(1,:) !write(*,*) 'TB22rad gamma33 = ',gammafi33mod !write(*,*) 'TB22rad eps76 = ',eps76(1,:) ENDDO ! ngrid ELSE ! if (strobel) !if strobel = false klw = 3.6E-4 ! fit strobel !eps_sw1Pa = 0.6 ! fit strobel eps_sw1Pa = 0.90 ! to fit NH temperature en 1D [ch4]=0.5 !eps_sw1Pa = 0.98 ! to fit NH temperature en 3D = fit 1D [ch4]=0.25 ksw = 8.5E-5*(0.9/(eps_sw1Pa-0.1)-1.) do l=1, nlayer do ig=1, ngrid rho(ig,l)=pplay(ig,l)/(r*pt(ig,l)) eps23(ig,l) = 0.1+ 0.9/(1.+ksw/rho(ig,l)) eps76(ig,l) = 1./(1.+klw/rho(ig,l)) end do end do END IF end subroutine nlte_ch4 real function k4zT(temp,vmr) implicit none real, intent(in) :: temp,vmr k4zT=1.9e-15*exp((temp-240.)/70.)*(1+10.*vmr) return end function k4zT !! Log log affine and square function for ratio adjustment and extrapolation real function log_square(x,y1,y2,x1,x2,coef) implicit none real, intent(in) :: x,y1,y2,x1,x2,coef real aa,bb aa=(log(y2)-log(y1))/((log(coef*x2))**4-(log(coef*x1)**4)) bb=log(y2)-aa*(log(coef*x2))**4 log_square=exp(aa*(log(coef*x))**4+bb) return end function log_square real function log_affine(x,y1,y2,x1,x2) real, intent(in) :: x,y1,y2,x1,x2 real aa,bb aa=(log(y2)-log(y1))/(log(x2)-log(x1)) bb=log(y2)-aa*log(x2) log_affine=exp(aa*log(x)+bb) return end function log_affine !! Ratio of gamma depending on T. Linear approximation real function ratio_affine(temp,n) implicit none real, intent(in) :: temp,n real aa,bb,ratio0 real log_affine aa=(1.-0.2)/(100.-40.) bb=1.-aa*100. ratio0=aa*temp+bb ratio_affine=log_affine(n,1.,ratio0,1e15,1e22) return end function ratio_affine !! Ratio of gamma depending on T. Linear approximation real function ratio_square(temp,n) implicit none real, intent(in) :: temp,n real aa,bb,ratio0,coef real log_square coef=1.e-12 aa=(1.-0.2)/(100.-40.) bb=1.-aa*100. ratio0=aa*temp+bb ratio_square=log_square(n,1.,ratio0,1e15,1e22,coef) return end function ratio_square