| 1 |
% \documentclass[12pt]{article} |
| 2 |
% \usepackage{amssymb} |
| 3 |
% |
| 4 |
% \usepackage{graphics} |
| 5 |
% |
| 6 |
% |
| 7 |
% \oddsidemargin -4mm \evensidemargin 0mm |
| 8 |
% \textwidth 165mm |
| 9 |
% \textheight 230mm |
| 10 |
% \topmargin -2mm \headsep -2mm |
| 11 |
% \renewcommand{\baselinestretch}{1.5} |
| 12 |
% \begin{document} |
| 13 |
% |
| 14 |
|
| 15 |
%%%--------------------------------------%%% |
| 16 |
\def\deg{$^o$} |
| 17 |
\section{Bulk Formula Package} |
| 18 |
|
| 19 |
author: Stephanie Dutkiewicz\\ |
| 20 |
|
| 21 |
\noindent |
| 22 |
Instead of forcing the model with heat and fresh water flux data, |
| 23 |
this package calculates these fluxes using the changing sea surface |
| 24 |
temperature. We need to read in some atmospheric data: |
| 25 |
{\bf air temperature, air humidity, down shortwave radiation, |
| 26 |
down longwave radiation, precipitation, wind speed}. |
| 27 |
The current setup also reads in {\bf wind stress}, but this |
| 28 |
can be changed so that the stresses are calculated from the |
| 29 |
wind speed. |
| 30 |
|
| 31 |
The current setup requires that there is the thermodynamic-seaice package |
| 32 |
({\it pkg/thsice}, also refered below as seaice) |
| 33 |
is also used. It would be useful though to have it also |
| 34 |
setup to run with some very simple parametrization of the sea ice. |
| 35 |
|
| 36 |
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
| 37 |
|
| 38 |
\vspace{1cm} |
| 39 |
|
| 40 |
\noindent |
| 41 |
The heat and fresh water fluxes are calculated in {\it bulkf\_forcing.F} |
| 42 |
called from {\it forward\_step.F}. These fluxes are used over open water, |
| 43 |
fluxes over seaice are recalculated in the sea-ice package. |
| 44 |
Before the call to {\it bulkf\_forcing.F} we call |
| 45 |
{\it bulkf\_fields\_load.F} to find the current atmospheric conditions. |
| 46 |
The only other changes to the model code come from the initializing |
| 47 |
and writing diagnostics of these fluxes. |
| 48 |
|
| 49 |
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
| 50 |
\vspace{1cm} |
| 51 |
\noindent |
| 52 |
{\bf \underline{subroutine BULKF\_FIELDS\_LOAD}} |
| 53 |
|
| 54 |
\noindent |
| 55 |
Here we find the atmospheric data needed for the bulk formula |
| 56 |
calculations. These are read in at periodic intervals and |
| 57 |
values are interpolated to the current time. The data file names |
| 58 |
come from {\bf data.blk}. The values that can be read in are: |
| 59 |
air temperature, air humidity, precipitation, |
| 60 |
down solar radiation, down long |
| 61 |
wave radiation, zonal and meridional wind speeds, total wind |
| 62 |
speed, net heat flux, net freshwater forcing, cloud cover, |
| 63 |
snow fall, zonal and meridional wind stresses, and SST and SSS |
| 64 |
used for relaxation terms. |
| 65 |
Not all these files are necessary or used. For instance cloud |
| 66 |
cover and snow fall are not used in the current bulk formula |
| 67 |
calculation. If total wind speed is not supplied, wind speed |
| 68 |
is calculate from the zonal and meridional components. If |
| 69 |
wind stresses are not read in, then the stresses are calculated |
| 70 |
from the wind speed. Net heat flux and net freshwater can be |
| 71 |
read in and used over open ocean instead of the bulk formula |
| 72 |
calculations (but over seaice the bulkf formula is always |
| 73 |
used). This is "hardwired" into {\it bulkf\_forcing} and |
| 74 |
the "ch" in the variable names suggests that this is "cheating". |
| 75 |
SST and SSS need to be read in if there is any relaxation used. |
| 76 |
|
| 77 |
|
| 78 |
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
| 79 |
|
| 80 |
|
| 81 |
\vspace{1cm} |
| 82 |
\noindent |
| 83 |
{\bf \underline{subroutine BULKF\_FORCING}} |
| 84 |
|
| 85 |
\noindent |
| 86 |
In {\it bulkf\_forcing.F}, we calculate heat and fresh water |
| 87 |
fluxes (and wind stress, if necessary) for each grid cell. |
| 88 |
First we determine if the grid cell is open water or seaice |
| 89 |
and this information is carried by {\bf iceornot}. There is |
| 90 |
a provision here for a different designation if there is |
| 91 |
snow cover (but currently this does not make any difference). |
| 92 |
We then call {\it bulkf\_formula\_lanl.F} which provides |
| 93 |
values for: up long wave radiation, latent and sensible heat |
| 94 |
fluxes, the derivative of these three with respect to surface |
| 95 |
temperature, wind stress, evaporation. |
| 96 |
Net long wave radiation is calculated from the combination |
| 97 |
of the down long wave read in and the up long wave calculated. |
| 98 |
|
| 99 |
We then find the albedo of the surface - with a call to |
| 100 |
{\it sfc\_albedo} if there is sea-ice (see the seaice package |
| 101 |
for information on the subroutine). If the grid cell is open |
| 102 |
ocean the albedo is set as 0.1. Note that this is a parameter |
| 103 |
that can be used to tune the results. The net short wave |
| 104 |
radiation is then the down shortwave radiation minus the |
| 105 |
amount reflected. |
| 106 |
|
| 107 |
If the wind stress needed to be calculated in {\it bulkf\_formula\_lanl.F}, |
| 108 |
it was calculated to grid cell center points, so in {\it bulkf\_forcing.F} |
| 109 |
we regrid to {\bf u} and {\bf v} points. We let the model know |
| 110 |
if it has read in stresses or calculated stresses by the switch |
| 111 |
{\bf readwindstress} which is can be set in data.blk, and defaults |
| 112 |
to {\bf .TRUE.}. |
| 113 |
|
| 114 |
We then calculate {\bf Qnet} and {\bf EmPmR} that will be used |
| 115 |
as the fluxes over the open ocean. There is a provision for |
| 116 |
using runoff. If we are "cheating" and using observed fluxes |
| 117 |
over the open ocean, then there is a provision here to |
| 118 |
use read in {\bf Qnet} and {\bf EmPmR}. |
| 119 |
|
| 120 |
The final call is to calculate averages of the terms found |
| 121 |
in this subroutine. |
| 122 |
|
| 123 |
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
| 124 |
\vspace{1cm} |
| 125 |
\noindent |
| 126 |
{\bf {\underline{ subroutine BULKF\_FORMULA\_LANL}}} |
| 127 |
|
| 128 |
\noindent |
| 129 |
This is the main program of the package where the |
| 130 |
heat fluxes and freshwater fluxes over ice and |
| 131 |
open water are calculated. Note that this subroutine |
| 132 |
is also called from the seaice package during the |
| 133 |
iterations to find the ice surface temperature. |
| 134 |
|
| 135 |
Latent heat ($L$) used in this subroutine |
| 136 |
depends on the state of the surface: vaporization for |
| 137 |
open water, fusion and vaporization for ice surfaces. |
| 138 |
Air temperature is converted from Celsius to Kelvin. |
| 139 |
If there is no wind speed ($u_s$) given, then the wind speed |
| 140 |
is calculated from the zonal and meridional components. |
| 141 |
|
| 142 |
We calculate the virtual temperature: |
| 143 |
\[ |
| 144 |
T_o = T_{air} (1+\gamma q_{air}) |
| 145 |
\] |
| 146 |
where $T_{air}$ is the air temperature at $h_T$, $q_{air}$ is |
| 147 |
humidity at $h_q$ and $\gamma$ is a constant. |
| 148 |
|
| 149 |
The saturated vapor pressure is calculate (QQ ref): |
| 150 |
\[ |
| 151 |
q_{sat} = \frac{a}{p_o} e^{L (b-\frac{c}{T_{srf}})} |
| 152 |
\] |
| 153 |
where $a,b,c$ are constants, $T_{srf}$ is surface temperature |
| 154 |
and $p_o$ is the surface pressure. |
| 155 |
|
| 156 |
The two values crucial for the bulk formula calculations are |
| 157 |
the difference between air at sea surface and sea surface temperature: |
| 158 |
\[ |
| 159 |
\Delta T = T_{air} - T_{srf} +\alpha h_T |
| 160 |
\] |
| 161 |
where $\alpha$ is adiabatic lapse rate and $h_T$ is the height |
| 162 |
where the air temperature was taken; and the difference |
| 163 |
between the air humidity and the saturated humidity |
| 164 |
\[ |
| 165 |
\Delta q = q_{air} - q_{sat}. |
| 166 |
\] |
| 167 |
|
| 168 |
We then calculate the turbulent exchange coefficients |
| 169 |
following Bryan et al (1996) and the numerical scheme |
| 170 |
of Hunke and Lipscombe (1998). |
| 171 |
We estimate initial values for the exchange coefficients, $c_u$, |
| 172 |
$c_T$ and $c_q$ as |
| 173 |
\[ |
| 174 |
\frac{\kappa}{ln(z_{ref}/z_{rou})} |
| 175 |
\] |
| 176 |
where $\kappa$ is the Von Karman constant, $z_{ref}$ is a |
| 177 |
reference height and $z_{rou}$ is a roughness length scale |
| 178 |
which could be a function of type of surface, but is here set |
| 179 |
as a constant. Turbulent scales are: |
| 180 |
\begin{eqnarray} |
| 181 |
u^* & = & c_u u_s \nonumber\\ |
| 182 |
T^* & = & c_T \Delta T \nonumber\\ |
| 183 |
q^* & = & c_q \Delta q \nonumber |
| 184 |
\end{eqnarray} |
| 185 |
|
| 186 |
We find the "integrated flux profile" for momentum and stability |
| 187 |
if there are stable QQ conditions ($\Upsilon>0$) : |
| 188 |
\[ |
| 189 |
\psi_m = \psi_s = -5 \Upsilon |
| 190 |
\] |
| 191 |
and for unstable QQ conditions ($\Upsilon<0$): |
| 192 |
\begin{eqnarray} |
| 193 |
\psi_m & = & 2 ln(0.5(1+\chi)) + ln(0.5(1+\chi^2)) - 2 \tan^{-1} \chi + \pi/2 |
| 194 |
\nonumber \\ |
| 195 |
\psi_s & = & 2 ln(0.5(1+\chi^2)) \nonumber |
| 196 |
\end{eqnarray} |
| 197 |
where |
| 198 |
\[ |
| 199 |
\Upsilon = \frac{\kappa g z_{ref}}{u^{*2}} (\frac{T^*}{T_o} + |
| 200 |
\frac{q^*}{1/\gamma + q_a}) |
| 201 |
\] |
| 202 |
and $\chi=(1-16\Upsilon)^{1/2}$. |
| 203 |
|
| 204 |
The coefficients are updated through 5 iterations as: |
| 205 |
\begin{eqnarray} |
| 206 |
c_u & = & \frac {\hat{c_u}}{1+\hat{c_u}(\lambda - \psi_m)/\kappa} \nonumber \\ |
| 207 |
c_T & = & \frac {\hat{c_T}}{1+\hat{c_T}(\lambda - \psi_s)/\kappa} \nonumber \\ |
| 208 |
c_q & = & c'_T |
| 209 |
\end{eqnarray} |
| 210 |
where $\lambda =ln(h_T/z_{ref})$. |
| 211 |
|
| 212 |
We can then find the bulk formula heat fluxes: |
| 213 |
|
| 214 |
\vspace{.2cm} |
| 215 |
\noindent |
| 216 |
Sensible heat flux: |
| 217 |
\[ |
| 218 |
Q_s=\rho_{air} c_{p_{air}} u_s c_u c_T \Delta T |
| 219 |
\] |
| 220 |
|
| 221 |
\vspace{.2cm} |
| 222 |
\noindent |
| 223 |
Latent heat flux: |
| 224 |
\[ |
| 225 |
Q_l=\rho_{air} L u_s c_u c_q \Delta q |
| 226 |
\] |
| 227 |
|
| 228 |
\vspace{.2cm} |
| 229 |
\noindent |
| 230 |
Up long wave radiation |
| 231 |
\[ |
| 232 |
Q_{lw}^{up}=\epsilon \sigma T_{srf}^4 |
| 233 |
\] |
| 234 |
where $\epsilon$ is emissivity (which can be different for |
| 235 |
open ocean, ice and snow), $\sigma$ is Stefan-Boltzman constant. |
| 236 |
|
| 237 |
We calculate the derivatives of the three above functions |
| 238 |
with respect to surface temperature |
| 239 |
\begin{eqnarray} |
| 240 |
\frac{dQ_s}{d_T} & = & \rho_{air} c_{p_{air}} u_s c_u c_T \nonumber \\ |
| 241 |
\frac{dQ_l}{d_T} & = & \frac{\rho_{air} L^2 u_s c_u c_q c}{T_{srf}^2} \nonumber \\ |
| 242 |
\frac{dQ_{]lw}^{up}}{d_T} & = & 4 \epsilon \sigma t_{srf}^3 \nonumber |
| 243 |
\end{eqnarray} |
| 244 |
|
| 245 |
And total derivative $\frac{dQ_o}{dT}= \frac{dQ_s}{dT} + |
| 246 |
\frac{dQ_l}{dT} + \frac{dQ_{lw}^{up}}{dT}$. |
| 247 |
|
| 248 |
|
| 249 |
|
| 250 |
|
| 251 |
If we do not read in the wind stress, it is calculated here. |
| 252 |
|
| 253 |
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
| 254 |
|
| 255 |
\vspace{1cm} |
| 256 |
|
| 257 |
\noindent |
| 258 |
{\bf {\underline{Initializing subroutines}}} |
| 259 |
|
| 260 |
\noindent |
| 261 |
{\it bulkf\_init.F}: |
| 262 |
Set bulkf variables to zero. |
| 263 |
|
| 264 |
\noindent |
| 265 |
{\it bulkf\_readparms.F}: |
| 266 |
Reads {\bf data.blk} |
| 267 |
|
| 268 |
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
| 269 |
|
| 270 |
\vspace{1cm} |
| 271 |
|
| 272 |
\noindent |
| 273 |
{\bf {\underline{Diagnostic subroutines}}} |
| 274 |
|
| 275 |
\noindent |
| 276 |
{\it bulkf\_ave.F}: |
| 277 |
Keeps track of means of the bulkf variables |
| 278 |
|
| 279 |
\noindent |
| 280 |
{\it bulkf\_diags.F}: |
| 281 |
Finds averages and writes out diagnostics |
| 282 |
|
| 283 |
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
| 284 |
\vspace{1cm} |
| 285 |
|
| 286 |
\noindent |
| 287 |
{\bf {\underline{Common Blocks}}} |
| 288 |
|
| 289 |
\noindent |
| 290 |
{\it BULKF.h}: BULKF Variables, data file names, and logicals |
| 291 |
{\bf readwindstress} and {\bf readsurface} |
| 292 |
|
| 293 |
\noindent |
| 294 |
{\it BULKF\_DIAGS.h}: matrices for diagnostics: averages of fields |
| 295 |
from {\it bulkf\_diags.F} |
| 296 |
|
| 297 |
\noindent |
| 298 |
{\it BULKF\_ICE\_CONSTANTS.h}: |
| 299 |
all the parameters need by the ice model and in the bulkf formula |
| 300 |
calculations. |
| 301 |
|
| 302 |
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
| 303 |
\vspace{1cm} |
| 304 |
|
| 305 |
\noindent |
| 306 |
{\bf {\underline{Input file DATA.ICE}}} |
| 307 |
|
| 308 |
\noindent |
| 309 |
We read in the file names of atmospheric data used in |
| 310 |
the bulk formula calculations. Here we can also set |
| 311 |
the logicals: {\bf readwindstress} if we read in the |
| 312 |
wind stress rather than calculate it from the wind |
| 313 |
speed; and {\bf readsurface} to read in the surface |
| 314 |
temperature and salinity if these will be used as |
| 315 |
part of a relaxing term. |
| 316 |
|
| 317 |
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
| 318 |
\vspace{1cm} |
| 319 |
|
| 320 |
\noindent |
| 321 |
{\bf {\underline{Important Notes}}} |
| 322 |
|
| 323 |
\noindent |
| 324 |
{\bf 1)} heat fluxes have different signs in the ocean and ice |
| 325 |
models. |
| 326 |
|
| 327 |
\noindent |
| 328 |
{\bf 2)} {\bf StartIceModel} must be changed in {\bf data.ice}: |
| 329 |
1 (if starting from no ice), 0 (if using pickup.ic file). |
| 330 |
|
| 331 |
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
| 332 |
|
| 333 |
\vspace{1cm} |
| 334 |
|
| 335 |
\noindent |
| 336 |
{\bf {\underline{References}}} |
| 337 |
|
| 338 |
|
| 339 |
\vspace{.2cm} |
| 340 |
|
| 341 |
\noindent |
| 342 |
Bryan F.O., B.G Kauffman, W.G. Large, P.R. Gent, 1996: |
| 343 |
The NCAR CSM flux coupler. Technical note TN-425+STR, |
| 344 |
NCAR. |
| 345 |
|
| 346 |
\vspace{.2cm} |
| 347 |
|
| 348 |
\noindent |
| 349 |
Hunke, E.C and W.H. Lipscomb, circa 2001: CICE: the Los Alamos |
| 350 |
Sea Ice Model Documentation and Software User's Manual. |
| 351 |
LACC-98-16v.2.\\ |
| 352 |
(note: this documentation is no longer available as CICE has progressed |
| 353 |
to a very different version 3) |
| 354 |
|
| 355 |
|
| 356 |
|
| 357 |
|
| 358 |
|
| 359 |
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% |
| 360 |
% \end{document} |
Parent Directory
| 
Revision Log
| 
Revision Graph