Accessing the data with LAS

1. Step-by-step example for Greenland annual mean surface temperature time series 21,000-0 BP.
   
   
1. Start: http://apdrc.soest.hawaii.edu/
2. select Data
3. select Model Results
4. select ECBilt-CLIO Transient climate simulation (LAS)
5. select 'SIM2bl' and then 'SIM2bl Annual atmosphere'
6. select ' Surface Temperature' and click on 'Next' 
7. (note only one variable can be processed with the current server)
8. select 'View mode - Time series', 'Output - Line Plot'
9. use the interactive map to click on Greenland 
   (or use the coordinate selector on the right for entering the coordinate)
10. select the time range: The first column is the index ranging from 0-20999
    and the right gives you the corresponding year with the unconventional
    counting from in years after 21,000 BP.
11. Using the default plotting options, submit the request to the LAS server by clicking on 'Next'
    (this process may take some seconds to minutes. Sometimes the server may return with a status message, where you can check the status of the request again)

2. Map of SST 11,000 BP

3. Creating a SST difference plot 21ka BP - 1ka BP
 

1. start LAS
2. select SIM2bl Annual ocean
3. select Sea Surface Temperature (SST) and click Next
   (make sure that only one variable is selected from the list)
4. On the left hand side, click on "Define Variable"
   (this opens extra options in the right window)
5. Select analysis type : average
6. enter a name for the new variable
7. click on the checkbox for the t-axis 
   (apply to these axes : T [ *])
8. select the time 0-1000 (e.g. 21ka BP - 20ka BP) and click Next
9. use the button Variables on the left hand side to return to the variable list
   (you should see your newly defined variable at the top of the list)
   
10. deselect your defined variable for first and select SST again (+Next)
11. repeat steps 4-9 with time 20000-20990 (+Next)
    NOTE if you cannot return to the variable list, go back to the variable list. You will see the defined variables 
12. select "compare two variables" from the left menu
13. click on "variable 1" in the left menu and select the defined 21ka SST field 
14. click on "variable 2" in the left menu and select the defined 1ka SST field
15. choose options for plot customization (e.g. palette anomaly) (+Next)
    

---

How to interpret the time axis?

The transient simulation started from an LGM equilibrium state with boundary conditions 21,000BP.
The timing of forced climate variability in the model simulation is dictated by the timing of the changes in the external forcing, i.e. the boundary conditions of the model.
Whereas the timing of the orbital forcing is exact, uncertainties in the greenhouse gas forcings are unavoidable (due to uncertainties in the depth-age conversion for ice cores and ice-age gas-ice differences). Uncertainties from the proxy records of sea level change (etc) that contrain the ICE4 icesheet reconstruction must also be considered.
Therefore, phase relationships to proxy data records must be considered within these uncertainties.
More details and discussion can be found in Timm and Timmermann (2007) and Timmermann et al. (submitted to Journal of Climate, 2008)




Annual mean data:

Atmospheric data:

For the interpretation of the transient climate simulation, we therefore interpret the time steps in the annual mean model output as:

model index	LAS time	OPeNDAP time	year BP
1	0	0	21,000
2	1	1	20,999
3	2	2	20,998
...	...	...	...
21000	20999	20999	1

Oceanic data:

To reduce the large data amounts of the 3-d ocean model, only every tenth year was saved.
The first annual mean output for the ocean variables corresponds to year 10 of the 21,000-year long simulation. The timescale for the oceanic transient climate simulation is therefore: 

model index	LAS time	OPeNDAP time	year BP
1	0	0	21,000
2	10	1	20,990
3	20	2	20,980
...	...	...	...
2100	20990	20990	10

Seasonal mean data:

Seasonal mean output is available for the atmospheric data only. Defining seasons in a model simulation with changing orbital parameters can be done in at least two different ways. Our model uses a 360-day calendar with an division of the year into four seasons of equal length (90 days). The vernal equinox is fixed to day 81 in the model. The days over which seasons are averaged are kept unchanged throughout the simulation. The resulting fixed calendar seasons are the output that is aviailable online. We refer to Timm et al. (2008)  for a more detailed discussion.
 


Atmospheric data:

For the interpretation of the transient climate simulation, we therefore interpret the time steps in the seasonal mean model output as:

model index	LAS time	OPeNDAP time	year BP
1	0.00	0.00	21,000 DJF
2	0.25	0.25	21,000 MAM
3	0.50	0.50	21,000 JJA
4	0.75	0.75	21,000 SON
5	1.00	1.00	20999 DJF
6	1.25	1.25	20999 MAM
...	...	...	...
83999	20999.50	20999.50	1 JJA
84000	20999.75	20999.75	1 SON

(you can convert from year BP to the LAS time interpretation here)

Note: The model's seasonal averaging routine produces undefined values in the first year every time the model is restarted. We used a restart interval of 1000 model years. Hence, the values at time steps (1,2,3,4) (4001,4002,4003,4004) (8001,8002,8003,8004) etc. are filled with NaN values.
 

---

Find the right LAS time index range for a year BP:

enter year BP

Atmospheric Data (annual)| Atmospheric Data (seasonal) | Oceanic Data (annual)

LAS year LAS Index