Please be aware that this is an experimental catalogue entry which uses an alternative source for the Land cover data designed for fast access long-timeseries for a single point, and it is not recommended for operational use.
The entry may be temporarily disabled or completely deprecated at any point and it does not come with the same level of operational support as the parent entry. Additionally, the formats and structure of the data files are subject to change.
Introduction
This dataset presented here is a subset of some parameters of the full CDS Land cover dataset on approximately 300m resolution and is stored in Analysis Ready Cloud Optimised (ARCO) format, which has been implemented for retrieving long time-series for a single point in a efficient way.
The dataset contains data from 1992 onwards. The version used to produce the data between 1992 to 2015 is v2.0.7cds and v2.1.1 after 2016.
Methodology
The process to generate this dataset involves:
- Fetching Satellite Land Cover data in netCDF format.
- Regridding from original tripolar coordinate system to a regular 0.25º longitude and latitude grid.
- Applying homogenization conventions, as applicable:
- Ensure latitude, longitude ranges are between [-90, +90], [-180, +180].
- Writing to an ARCO Zarr archive (which is similar in structure to netCDF).
- The CDS requests the data from the Zarr archive using xarray and writes the data to netCDF or CSV (as requested).
Data organization and access
The ARCO data is stored as Zarr datacubes.
The data is now available from the Climate Data Store (CDS), either interactively through its download web form or programmatically using the CDS API service:
Land cover classification time-series data from 1992 to present derived from satellite observations <link>
Spatial grid
The parameters have longitude and latitude dimensions and a resolution of 300m.
Please be aware when you select a "Location" on the form, the latitude and longitude values are rounded to the closest neighborhood point of the degrees grid. Is this true for this dataset? Could you explain what happen with the area selection → (Marc) These details depend on the client application that reads Zarr data and how it deals with subsetting, whether it does any kind of interpolation or not. My assumption is that for an area selection would happen exactly the same as for a single location selection, the closest neighborhood corners of the area polygon would be selected. Please confirm with Eduardo Damasio, and in case of doubt we can check together the code that consumes the Zarr file to confirm this.
Temporal frequency
Temporal frequency of the data is yearly but with one year delay to present.
Data format
In the CDS, there are the options of retrieving the data in netCDF format or CSV format.
Data update frequency
ORAS5 is updated yearly.
Variations in delivery times may occur due to the non-operational nature of this CDS service, as issues may arise which cause delays.
Parameters listings
Table 1 below lists the parameters available.
Table 1: Parameters
| Name | Variable name | Notes | |
|---|---|---|---|
| 1 | Change count | change_count | |
| 2 | Current pixel state | current_pixel_state | |
| 3 | Land cover class | land_cover_class | |
| 4 | Observation count | observation_count | |
| 5 | Processed flag | processed_flag |
Known issues
Please refer to the Land cover documentation.
