The Time Series tab on the Object Properties dialog box. is used for specifying MODFLOW 6 Time Series. It differs from time series defined in the Time Series dialog box in that formulas can be used to specify individual values in the series. These time series can be used for many transient data in MODFLOW 6 such as well pumping rates. The formulas in the time series will be evaluated at each location at which the time series is used when exporting the MODFLOW 6 input files and will result in a separate time series for each location. Thus use of these "dynamic" time series can result in a large number of time series being defined in the MODFLOW 6 input files. Limited testing with MODFLOW 6 version 6.4.1 indicates that with 10,000 time series, MODFLOW 6 execution time is not greatly affected but that with 100,000 time series MODFLOW 6 execution time is slowed substantially. The use of "static" time series defined in the Time Series dialog box does not result in the proliferation of times series in the MODFLOW input files. Thus, static time series do not affect execution time substantially when used at multiple locations.
With some boundary conditions, ModelMuse may modify the user-specified formula when the formula is evaluated. For example, with well pumping rates, the formulas can be "Calculated", "Direct", or "Total per Layer.". With "Calculated" and "Total per Layer", the user specified formula is ordinarily modified. However, this modification can not be used directly with time series. Instead, you can specify a formula for the scale factor of the time series or if the boundary allows "multipliers" to be used, the multiplier can be used to replicate the effect of the changes in the formula that would ordinarily be applied by ModelMuse.
A video about this topic is available at https://www.usgs.gov/media/videos/using-modflow-6-times-series-modelmuse.
A time series is an alternative method in MODFLOW 6 for specifying time-varying data. The modeler specifies a series of times along with a corresponding series of values. The times do not need to correspond to the beginning and ending times of any stress period. Instead, MODFLOW will interpolate the values in the time series to determine the value applied to any time step. Three different interpolation methods can be used: STEPWISE, LINEAR, and LINEAREND.
•With STEPWISE, the values change abruptly at each new time and the value that is applied to a time step is a weighted average of the values for a time step.
•With LINEAR, values are interpolated linearly among the times and the value applied to any time step is a weighted average of the values for that time step.
•With LINEAREND, the values are also interpolated linearly but the value applied is the value for the end of the time step.
The times must be in ascending order. The value 3E30 can be used to indicate missing values in a time series.
For example, suppose that the following time series was specified as the well pumping rate in which the time steps each had a length of 1 and the starting time of the first time step was zero.
Time |
Values |
0 |
-1 |
0.5 |
-2 |
2 |
-3 |
If STEPWISE was used, the value applied for the first time step would be ‑1.5 for the first time step and ‑2 for the second time step. If LINEAR was used, the value applied for the first time step would be ‑1.8333 for the first time step and ‑2.6667 for the second time step. If LINEAREND was used, the value applied for the first time step would be ‑2.3333 for the first time step and ‑3 for the second time step.
An example of when STEPWISE might be appropriate would be for a well pumping rate in which the pumping rate is changed abruptly from time to time but in which there is no gradual change. An example of when LINEAR might be appropriate would be for representing seasonal change in the maximum evapotranspiration rate. An example of when LINEAREND might be appropriate would be for a specified head boundary such as a lake where the lake level changes gradually.
Time series can also have a scale factor. MODFLOW will modify all the values in a time series by multiplying them by the scale factor. PEST parameters can be used with time series. The PEST parameter will modify the scale factor for the time series. The scale factor can be a formula for dynamic time series define for an object.
In ModelMuse, certain time varying values can depend on the length or area of intersection of the object with a cell. However, this does not apply when a time series is used. In such cases, the time series is used without modification. Similarly, if a PEST parameter is applied in the Object Properties dialog box, it will not apply to any time series. Instead only PEST parameters in the Time Series Dialog box or the Time Series tab of the Object Properties dialog box will apply to a time series.
