First, to download the latest codes from

   http://oceans.deas.harvard.edu/HOPS/

Download upgrades (upgraded biology; provided-Orlanski boundary conditions;
direct input of velocity fields) from

   http://oceans.deas.harvard.edu/haley/Nello/

Grab the files

   pe_nello.tar.gz  -  Upgraded PE routines only
   pi_nello.tar.gz  -  Upgraded PE_Initial routines only

These are only the routines affected by the upgrades.  They simply
replace the corresponding routines from the HOPS web site.

The simplest way to read in the biological data is to use the C-Preprocessing
option "trcascii".  This option assumes that you have the biological fields
mapped onto the same grid as the physical OA fields (i.e. same horizontal 2D
grid and same constant depths).  The data is then assumed to be a simple ascii
file structured to be read as

      do 20 n = 3,nt   
         do 30 i=1,im
            do 30 j=1,jm
               read(trcinp,*)(t(i+(j-1)*im,level,n),level=1,kfld)
 30      continue
 20   continue

where i varies in the x direction, j in the y direction, level in the
vertical and n is the tracer number (the ordering of tracers is implied
in the ordering of the variable "tname" in blkdat.F)

A student of Allan's (Patrica Moreno) has provided me more detailed
instructions on creating biological initializations using a modified
Dusenberry biological model.

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
HOPS: Initialization of biological fields

Assumptions:
   Observed Variables:  Nitrate      NO3
                        Ammonium     NH4
                        Chlorophyll  Chl (or fluorescence)

   Non-observed Variables:  Phytoplankton  P
                            Zooplankton    Z
                            Detritus       D

(1) Objectively analysize the observed biological variables

       Create MODS format data files in which NO3, NH4 and Chl are
       stored as temperature or salinity

(2) Create corresponding 3D fields for non-observed variables

       set P proportional to Chl
       set Z proportional to P
       set D proportional to Z

       (Note:  In the past Pierre has had some success setting
               Z and D to satisfy the steady-state equations.
               That did not work so well for this example.)

       use the matlab script

          bio_correlation_sol.m

          must include in bio_correlation_sol.m:
             name of OA files
             biological parameters used in PE model

(3) Save initial biological fields in ASCII format used by PE_initial

       At the end of bio_circadian_sol.m, bio_save.m is called which
       saves the biological variables in the ASCII format used by PE_initial

       Order:  NO3, Pno3, Z, NH4, D, Chl, Pnh4
               where
                 Pno3 = phytoplankton biomasss due to nitrate uptake
                 Pnh4 = phytoplankton biomass due to ammonium uptake
                 Total phytoplankton = Pno3 + Pnh4

       Prior to saving, the variables must be reshaped:
          Ex:  [m n o] = size(NO3);
               NO3 = reshape (NO3,m*n,o);

       the output of this is bio_fields.dat

       Hint:  Save a copy of bio_circadian_sol.m together with
              bio_fields.dat to keep track of what data and
              biological variables were used to calculate
              bio_fields.dat

(4) Process through PE_initial

       Compile pe_initial with biology option(s)

# -Dtrcascii       ascii file input for biological tracers
# -DbioDuse        Dusenberry biological model (if desired)
# -Dreadbioerr     read error fields for biological tracers (if desired)
                   (technically, this last is for assimilation only)
                   if readbioerr is used, error fields must be present
                   in bio_fields.dat in the order
                   (for each time)
                   NO3, Pno3, Z, NH4, D, Chl, Pnh4,
                   NO3 error, Pno3 error, Z error, NH4 error, D error,
                   Chl error, Pnh4 error

       Changes to pi_ini.in
          NT = 9  number of tracers
          ITRC = 1  Read additional tracers
          7th line of filenames: bio_fields.dat

          If pe_initial is compiled with bioduse option, there is
          no need to set IBIOM.

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
additional notes on the Dusenberry model:

A definition of "cellular NH4" (qn4) and "cellular NO3" (qn3)
might come in handy.

In HOPS:

p  = total phytoplankton = qn3 + qn4

qn4 = fraction of phytoplankton due to recycled production
qn3 = fraction of phytoplankton due to new production

Let

Q = (NO3/KNO3+NO3)*exp(-NO3inh*NH4) + (NH4/KNH4+NH4)

then

qn3 = (NO3/KNO3+NO3)*exp(-NO3inh*NH4)/Q*P
qn4 = (NH4/KNH4+NH4)/Q*P

where

KNO3   = half-sturation constant for NO3
KNH4   = half-saturation constant for NH4
NO3inh = NH4 inhibition parameter