GPKG Time Series Format Specification

The below section describe the TUFLOW GPKG time series format specification. The specification sits ontop of the GPKG v1.2 Specification with two additional required tables. The additional tables and details of the GIS layers are descibed below with examples. The examples are based a the 1D/2D linked TUFLOW-SWMM example model based on TUFLOW's module 6 tutorial model - M06_5m_003.tcf.

TUFLOW Time Series Version

An additional table called TUFLOW_timeseries_version should be present in the GPKG. This table contains the version number of the TUFLOW GPKG time series format and should be set to 1.0.0. The table should have the following columns:

Column Name	Type	Description
Version	TEXT	Major.Minor.Patch version number

Time Series Info

An additional table called Timeseries_info should be present in the GPKG. The purpose of this table is to list the available result types and relevant information for each type. The table should have the following columns:

Column Name	Type	Description
row	INTEGER	Primary key
Table_name	TEXT	Reference to the GIS table that the result sits within
Count	INTEGER	Number of unique elements
Reference_time	TEXT	Reference time in the form of: [UNITS] since YYYY-MM-DD hh:mm:ss
dt	FLOAT	Output timestep (units defined in Reference_time)
Column_name	TEXT	Name of the column in the GIS table containing the result
Series_name	TEXT	Name of the result type
Series_units	TEXT	Result type units

GIS Layers

The results are stored in GPKG layers, typically between 1 - 3 layers are stored in the result file depending on the output (one for point results, one for line results, one for polygon results). There is no requirement against storing results from multiple model runs in a single GPKG, this is just not typically how outputs are written.

The column names of the GPKG layers should follow the convention below:

Column Name	Type	Description
fid	INTEGER	Primary key
geometry	GEOMETRY	GeoPackage Geometry
ID	TEXT	Element ID (e.g. Pipe1)
Type	TEXT	Element Type (e.g. CONDUIT)
Source	TEXT	Result Source (e.g. SWMM)
TimeId	INTEGER	Unique id for each timestep
Time_relative	FLOAT	Output relative time (reference time and units in Timeseries_info)
Datetime	DATETIME	Output time in datetime
...[column for each result type for that given geometry type as listed in Timeseries_info- examples below]
Flow	FLOAT	Result flow
Velocity	FLOAT	Result velocity