Everyone should verify that their code is working properly by conducting
identical twin experiments as follows:


1. Run driver < input

2. Copy the output files to the input data files in the forcing directory:

cp output/EqPac/*.out forcing/EqPac/.
cp output/AS/*.out forcing/AS/.

3. Add the number of data points (listed at the top of the *.dat files) to
  the top of each *.out file in both the forcing/EqPac and the forcing/AS
  directories 

4. Save the old 'real' data files:

mv forcing/AS/AS_P.dat forcing/AS/AS_P.datreal
mv forcing/EqPac/EqPac_P.dat forcing/EqPac/EqPac_P.datreal
mv forcing/AS/AS_Z.dat forcing/AS/AS_Z.datreal
and so on.... for the other seven files

5. Move the *.out files to *.datITE (ITE = identical twin experiments)

mv forcing/AS/AS_P.out forcing/AS/AS_P.datITE
mv forcing/EqPac/EqPac_P.out forcing/EqPac/EqPac_P.datITE
mv forcing/AS/AS_Z.out forcing/AS/AS_Z.datITE
etc...

6. Copy the *.datITE files to *.dat files

cp forcing/AS/AS_P.datITE forcing/AS/AS_P.dat
cp forcing/EqPac/EqPac_P.datITE forcing/EqPac/EqPac_P.dat
cp forcing/AS/AS_Z.datITE forcing/AS/AS_Z.dat
etc...

7. Run driver < input  (Cost function should be less than 1.e-3)

8. Modify initial guesses of parameters in eco_common.f90 by 10-50%
  (Make sure to only modify the parameters which are being optimized!)

9. Run driver < input to make sure that a finite cost function is obtained

10. Run adjoint_driver < input to make sure that you get a cost function < 1e-3

11. Copy real data files back to *.dat files:

mv forcing/AS/AS_P.datreal forcing/AS/AS_P.dat
mv forcing/EqPac/EqPac_P.datreal forcing/EqPac/EqPac_P.dat
mv forcing/AS/AS_Z.datreal forcing/AS/AS_Z.dat
etc....