Input Data¶
Purpose & Scope¶
This section provides the normative specification of the input data model for RRINGG. It strictly describes input datasets — including supported formats, required metadata, and conversion rules.
Two primary data providers are supported:
Other sources may be used only if the datasets strictly comply with the formats and metadata requirements described below.
Key Principles¶
The required input files depend on the workflow selected by the user:
InSAR velocity field correction Primary input: - Line-of-Sight (LOS) mean velocity field raster - LOS vectors Optional input: - Station CSV files (GNSS time series or precomputed velocities)
Supported formats:
FormaTerre/FLATSIM products: - MVLOS_geo (Mean velocity raster, GeoTIFF) - CosNEU_geo (LOS vector map, GeoTIFF) Both are compliant with EPOS TCS conventions.
EPOS TCS DTSLOS: Line-of-sight displacement time series (CSV format)
I — FormaTerre/FLATSIM products¶
1. Data Description¶
The following products are required by RRINGG. For a detailed description, see the official FLATSIM documentation.
Products
Mean velocity (GeoTIFF, multi-band) - Purpose: mean LOS velocity field for correction workflows. - Data type:
float32orfloat64(units must be declared in metadata).Time Series (GeoTIFF, multi-band) - Purpose: per-pixel LOS displacement values over time. - Requirement: band order must match the
acquisition_datesarray in metadata.LOS vectors (GeoTIFF, 3-band) - Purpose: LOS unit vector per pixel. - Format: 3 bands corresponding to (vx, vy, vz) in the declared coordinate basis (see conventions).
Lookup tables (lut_radar, lut_ground) (GeoTIFF) - Purpose: mapping between radar and ground geometries. - Format: GeoTIFF arrays defining coordinate transformation (provider-specific conventions).
Optional station/velocity files (CSV) - Purpose: GNSS time series or velocity datasets for fitting/validation.
General Requirements
Raster format: GeoTIFF
Recommended compression:
LZWPreferred data type:
float32Explicit NoData value (e.g.,
999.0); do not mix NaN and sentinel values.
Mandatory Metadata Fields
Each product must include metadata (embedded or as a side-car .meta file).
Minimal schema:
DDSS_ID: string # e.g., LOS_DISPLACEMENT_TIMESERIES, MEAN_LOS_VELOCITY
Title: string # e.g., Time_Serie, velocity
User_ID: string # e.g., flatsim, CNRIREA
List_of_dates: [YYYY-MM-DDTHH:MM:SSZ] # required for time series, matches band order
Value_unit: string # e.g., cm/yr, rad/year
Software_version: string # e.g., NSBAS:v2.2.10
Applied_filter: string # e.g., Goldstein_0.50
Applied_corrections: string # e.g., ERAW:ERA5, No_Corrections
Service_used_for_generation: string
Geographic_CS_type_code: string # e.g., EPSG4326, RADAR_GEOMETRY
License: string or URI # usage restrictions
Used_DEM: string # e.g., SRTM_1arcsec
Band_description: string # description of bands
Product_url: string # download URL
Preview_url: string # preview image URL
2. Data Access¶
Go to the FLATSIM portal.
Scroll down to the zone map and select your region of interest (e.g., Balkans).
Log in or create an account (preferably via ORCID).
In the left panel, navigate to Parameters to filter products (e.g.,
orbitDirection=DESCENDING,relativeOrbitNumber=007).For RRINGG, download products with types AUXILIARYDATA and TIMESERIE (~20 GB total).
An example of the download workflow is shown below:
Note: If the download button is greyed out, the dataset may be under embargo. In that case: - A message “Restricted Access” will appear. - You must acknowledge the terms of use and, if necessary, contact the provider to request access.
Full product descriptions are available in the FLATSIM documentation.
II — EPOS TCS DTSLOS¶
1. Data Description¶
EPOS TCS provides InSAR time series in CSV format. These datasets require preprocessing before being ingested by RRINGG. For details, see the official EPOS documentation.
Products
InSAR product (DTSLOS) (CSV) - Content: displacement time series, mean velocity, topography, LOS unit vectors, spatial coherence. - Data type:
float32orfloat64.Metadata file (text) - Similar in structure to FLATSIM metadata. - Must include acquisition dates in chronological order.
CSV Fields
ID(string)Lat(decimal degrees)Lon(decimal degrees)Topo(m)Vel(cm/yr)Coer(unitless coherence)cosN, cosE, cosU(LOS vector components, unitless)TS(cm displacement time series)
The Displacement Time Series file structure should resemble the following:
The DTSLOS file is an ASCII CSV table, where each row represents a measurement point.
Mandatory Metadata Example
DDSS_ID: MEAN_LOS_VELOCITY
Title: Time_Serie
User_ID: flatsim
List_of_dates: [2015-05-07T12:25:18Z, 2015-05-31T12:25:20Z, ...]
Value_unit: rad/year
Software_version: NSBAS:v2.2.10
Applied_filter: sliding window average
Applied_corrections: ERAW:ERA5
Service_used_for_generation: WP12-DDSS-024 GDM-SAR Sentinel-1
Geographic_CS_type_code: EPSG4326
License: https://creativecommons.org/licenses/by-nc/4.0/
Used_DEM: SRTM V3, 1 arcsec
Band_description: LOS velocity [rad/yr], Radar illumination on DEM
Product_url: <download URL>
Preview_url: <preview URL>
2. Data Access¶
Visit the EPOS ICS portal.
Navigate: Data → Satellite Data → LOS Displacement Time Series.
On the map, select a region of interest.
In the popup, scroll down and click Product download (e.g.,
DTSLOS_CNRIREA_20160714_20231030_ZII6).Extract the downloaded archive to access CSV and metadata files.
An example of the download workflow is shown below:
III — Other Sources¶
Other providers may be used only if data are converted into the formats and conventions described above.
1. Naming Conventions¶
Recommended patterns:
[AGENCY]_DTS_[GEOMETRY]_[LOOKS]rlks.tiff
[AGENCY]_DTS_[GEOMETRY]_[LOOKS]rlks.meta
[AGENCY]_MVLOS_[GEOMETRY]_[LOOKS]rlks.tiff
[AGENCY]_MVLOS_[GEOMETRY]_[LOOKS]rlks.meta
[AGENCY]_Cos[COORD]_[GEOMETRY]_[LOOKS]rlks.tiff
[AGENCY]_LUT_[GEOMETRY]_[LOOKS]rlks.tiff
Components
AGENCY: data provider (e.g., CNES, CNRIREA)DTS: LOS displacement time seriesMVLOS: Mean LOS velocity rasterCos[COORD]: LOS vector raster (ENU or NEU convention)LUT: Look-up table raster for geometry conversionGEOMETRY:geo(ground, lat/lon) orradar(radar geometry)LOOKS: multi-looking factor in azimuth (e.g.,10rlks= averaged over 10 looks)
Best Practices
Always provide a
.metafile for each raster.Ensure metadata includes acquisition parameters and provenance.
Apply consistent naming across workflows to enable automation and reproducibility.
Example Table
File name |
Product type |
Description |
|---|---|---|
|
Time series |
LOS displacement raster, ground geometry, averaged over 10 looks. |
|
Metadata |
Metadata for mean velocity raster, ground geometry. |
|
LOS vectors |
ENU LOS vector raster in radar geometry. |
|
LUT |
Geometry conversion raster, radar geometry, 20 looks. |
2. Conventions¶
LOS Vectors
Must be unit vectors (vx, vy, vz).
Allowed conventions: - NEU (North, East, Up) - ENU (East, North, Up)
Convention must be explicitly declared in metadata.
Time-Series Band Ordering
Bands must be in strictly increasing chronological order.
acquisition_datesin metadata must match band count and useYYYY-MM-DDformat.