Concept - Data Representation
Examples¶
- Univariate Time Series
- Multivariate Time Series
- Univariate SpatioTemporal Series
- Multivariate SpatioTemporal Series
- Coupled Multivariate SpatioTemporal Series
Coordinates¶
Spatial¶
[Radius, Longitude, Latitude] [Altitude/Depth, Longitude, Latitude] [Channel, X, Y] Temporal¶
Variable¶
- Remote Sensing - Spectral Signature
- Ocean - Temperature, Height, Salinity, Colour
- Weather Station - Temperature, Precipitation, Wind Speed
Samples¶
- Ensembles/Realizations
- Patches
Dataset Structures¶
Univariate Time Series¶
D={tn,yn}n=1N,N=NTyn∈RDytn∈R+ Measurements:Time Stamps:yt∈RNT∈RNTyt∈Rtn∈R+
Shape¶
Irregular.
Example:
- Extreme Events
- Faulty Station
- ARGO Floats
Regular.
Examples:
- Single Weather Station - Max Temperature, Mean Temperature, Precipitation Accumulation
- Global Mean Surface Temperature Anomaly
Multivariate Time Series¶
D={tn,yn}n=1N,N=NTyn∈RDytn∈R+ Measurements:Time Stamps:Yt∈RNT×Dy∈RNTyn∈Rtn∈R+
Univariate SpatioTemporal Series¶
Coordinate-Based Representation¶
D={(tn,sn),yn}n=1Nyn∈RDytn∈R+ Measurements:Time Stamps:Spatial Coordinates:YtS∈RNT×Dy∈RNT∈RNT×Dsyn∈RDytn∈R+sn∈RDs Examples:
Field-Based Representation¶
D={tn,yn}n=1N,N=NTD=DyDΩyn∈RD Measurements:Time Stamps:Yt∈RNT×D∈RNTyn∈RDtn∈R+ Examples:
- Gridded Weather Station Data Product
- Sea Surface Height
- Sea Surface Temperature