NEA-1077 TIME2.

1. NAME OR DESIGNATION OF PROGRAM:  TIME2.

3. DESCRIPTION OF PROGRAM OR FUNCTION

TIME2 simulates the long-term evolution burial facilities for radioactive waste disposal, and forms part of a development programme in time-dependent probabilistic risk analysis. Environmental change in the model is essentially driven by changes in climatic conditions based on the geological record of climate changes and the astronomical or "Milankovitch" theory of climates. The environment is modelled up to the first glaciation from the present. Processes brought on by climatic change such as glacial advance, sea-level changes, water budget, and erosion effects are considered, as well as human intrusion effects, each as a separate sub-model within the overall framework of the model.

4. METHOD OF SOLUTION

TIME2 is a Monte Carlo simulation code which models environmental change by random sampling of the parameters which are anticipated to vary with time within quantifiable limits.  Probabilistic distribution functions for these parameters are provided by the user as input, and values are then sampled during each simulatiom of the code. Sampling occurs several times within each defined climate state type into which the changing climatic conditions from the present to the next Ice Age are divided (Temperate, Boreal, Tundra and Glacial). In addition a Savannah climate is modelled in the near future (500-1500 years) if the "greenhouse effect" is to be simulated). The sampling of these environmental paramaters occurs at specific time steps during a simulation which correspond either to the times of transition from one climate state to another, or to fixed user-specified times. The  climate state durations are themselves sampled at the start of each  simulation. Occurences of human intrusion events are sampled separately for each time step and their effects assessed. A particular case involving input values which simulate conditions for example for a given site and repository, would typically include several thousand simulations so as to simulate a range of possible outcomes resulting from the combined contributions of the many different sampled parameters.
The effects of many of the physical processes considered are determined using a set of geographic points selected to represent key points or regions of the study area. The changes in sea-level and point elevations caused by an advancing glacier are represented  using a "glacial picture" which maps their change with time as the glacial front approaches the site.

5. RESTRICTIONS ON THE COMPLEXITY OF THE PROBLEM

Number of fixed user-specified time steps : 20
Number of bin frequencies for output p.d.f.s at user-specified time  steps can be produced : 20
Number of parameters for which output p.d.f.s at user-specified time steps can be produced : 20
Maximum number of input pdfs : 200
Maximum number of bins in input pdfs : 50
Maximum number of geographic point : 50
Maximum number of system state (global) parameters per sub-model/section of the code : 20
Maximum number of times at which values in glacial pictures known :
100
Typically, many simulations (1000's) need to ne included within a case to obtain converged results.

6. TYPICAL RUNNING TIME

Run time for a TIME2 case depends on the computer system configuration and the TIME2 data specification. Run  times increase markedly with the quantity of output which is specified by the user in the T2CASE.DAT file, since a considerable amount of data has to be written to the disk. We recommend that minimum output be specified where possible.
For a Microvax II computer, with 5Mb of RAM memory, cases with minimum output have been executed in the following elapsed time:

No of simulations      CPU minutes      Elapsed minutes
-----------------      -----------      ---------------
      100                  7.4                7.7
     1000                 53.5               53.5
    10000                529.9              530.3

These times can be improved by compiling the code using the optimise and bo-degub options.

NEA-DB ran the test case included in this package on a  DEC VAX 8810 computer in 9 minutes of CPU time.

7. UNUSUAL FEATURES OF THE PROGRAM

The TIME2 model is suitable for the simulation of :
A shallow, engineered/non-engineered land-based repository;
Location of site in Britain (excluding N. Scotland and Ireland);
Timescales up to the next glacial maximum (c. 25,000 years), when  it is assumed that a shallow repository would be completely eroded;  Process modelled (each by individual sub-model called by the TIME2   executive:
    CLIMATE - selects climatic paramaters;
GLACIER - models ice sheet advance and its effects on topography  SEA LEVEL - models net changes in sea level, due to the greenhouse                effect and ice sheet growth;
GWATER - models land use change, surface water run-off,               infiltration and recharge;
DENUDE - models human intrusion and its effects on repository  status, and the net effects of all mechanisms which               disturb the repository and remove material from it.
Processes or events
       not modelled   Ice sheet erosion;
                      Wind erosion;
                      Landslips/flows/slides;
                      Meteorite (or other projectile) impact;
                      War or civil disturbance;
                      Flooding.

8. RELATED AND AUXILIARY PROGRAMS

SOURCE OF MODEL:
Framework of model is based on the ONWI Far-Field State Model produced by INTERA Inc. for U.S. Dept. of Energy. The individual sub-models which perform the technical processing are new.
Processing of Output:
A graphics post-processor exists though is not supplied, which provides, for parameters selected by the user in the input data, graphical presentations of their output pdf's at selected times, or  their mean/median values with distribution percentiles over the simulated time. This post-processor is called GRAFT2 and requires the SIMPLEPLOT (produced by Bradford University Software Services Limited) library of graphics routines. Other graphic routines could  potentially be used to display output data contained in the file T4GOUT.DAT.

10. REFERENCES:

- Dames and Moore:
  Design of TIME2 Code: Time Dependent Effects on Land 2 Type
  Repositories for Department of the Environment.
  DOE/RW/85/076 (July 1985).
- Dames and Moore:
  Test Report, TIME2 Version 1.0/1.2
  DOE/RW/88.054 (April 1988).
- Dames and Moore:
  Programmer's Manual, TIME2 Version 1.2
  DOE/RW/88.086 (April 1988).
- Dames and Moore:
  User Guide, TIME2 Version 1.2
  DOE/RW/88.088 (April 1988).

11. MACHINE REQUIREMENTS

   Memory necessary LE 5 Mn R.A.M.
   Hard Disk Capacity LE 430 mb.
Above are the dimensions of the code developer's system; much less is probably required by the model in practice.
High-Density Tape drive is used for archiving data files from cases  (output file containing detailed output of data is very large).
Input and Output data are held in formatted files on the hard disk,  and are not automatically printed out.
A system printer is desirable for printing input and output files for user inspection.
An HPGL-compatible plotter or other emulation which can be used with SIMPLEPLOT is needed for plotting output pdf's using GRAFT2, which calls on the SIMPLEPLOT library of routines, should GRAFT2 be requested.

13. OPERATING SYSTEM UNDER WHICH PROGRAM IS EXECUTED:  MicroVMS version  4.2.

VMS 5.1-1 (VAX 8810).

14. OTHER PROGRAMMING OR OPERATING INFORMATION OR RESTRICTIONS

Coding
is written to Digital Fortran 77 standard, though no Digital intrinsic functions are believed to be used.

15. NAME AND ESTABLISHMENT OF AUTHORS

Framework from ONWI, F.F.S.M. (Intera, 1983)

Technical sub-models; Dames & Moore 1987.

File name	File description	Records
NEA1077_01.001	Information file	87
NEA1077_01.002	TIME-2 FORTRAN-77 source	12635
NEA1077_01.003	VAX FORTRAN include file	63
NEA1077_01.004	Case control data	98
NEA1077_01.005	Site data	1015
NEA1077_01.006	Error message data	84
NEA1077_01.007	Output data for printing and examination	2362
NEA1077_01.008	Output to be used for GRAFT2 post-processor	1484
NEA1077_01.009	Output data on the mean and 95% confidence	23