MOLES v3 - Basic Formalism

( The basic formalism of MOLES-v3, no matter of standardisation )

A typical sequence of data capturing involves one or more projects under which a number of activities are undertaken, using appropriate tools and methods to produce the datasets. Following this typical sequence, the relevant metadata can be partitioned into the following main sections – helpful in mapping onto the most suitable standards from the ISO 19100 series.

  • Project section
  • Observation section (metadata regarding the methods used to obtained the data, the spatial and temporal sampling regime, quality etc.)
  • Observation collection section

1. Project Section

All the relevant metadata regarding the project under which the data is obtained

2. Observation Section

Observation act

According to [Fowler, 1998], an observation is an act associated with a discrete time instant or period which potentially can lead to a result, obtained using a specified procedure applied in-situ, remotely or ex-situ with respect to the sampling location. The result, the realisation of the observation objective, is an estimate of some property of an identifiable object, which is called feature-of-interest of the observation.
EXAMPLE: measurement of the rain rate at a specific location at 16:00 GMT. The feature-of-interest is the rain, the property is the rain rate, the procedure is a tipping bucket rain-gauge and the result is 40mm/hr.

 Here you can find a detailed document describing the concepts of the observation act; the observation procedure and its partitioning into Acquisition, Conditioning and Computation Process Steps; the property values and estimates and their domain.

Sampling Feature

ISO 19156 introduces the concept of sampling-feature to model the spatial aspects of sampling regime which enhances the concept of observation, making it more generic. A sampling-feature can be regarded as a subset of a feature-of-interest (set). Estimates of property values refer to a sampling-feature explicitly and to the corresponding feature-of-interest, which might not even be strictly defined, implicitly. A sampling-feature can be the target for an infinite number of observations.
EXAMPLE:measurements of air temperature or precipitation - the atmosphere can be the feature-of-interest and a specific point, a column or a flight line can be the sampling-feature.

Summarising, sampling-feature represents the geospatial coordinates of the observation result (the estimates of the property values) whereas the geospatial coordinates of the property values are the spatial extent of the feature-of-interest.Sampling-feature is the spatial component of the domain Dpro ( see this document)

NOTE: This differs from what is stated in ISO 19156/N2860 (section a common role for the spatial sampling feature is to host instruments or procedures deployed repetitevely or permanently. SF_SpatialSamplingFeature supports the association Platform which, if present, links the SF_SpatialSamplingFeature to an OM_Process deployed at it. The OM_Process has the role hostedProcedure with respect to the sampling-feature. On the contary, we claim that the location of observation procedure and sampling-feature are totally different concepts despite there are cases where location properties of a sampling-feature and observation procedure have the same values (i.e. measurements where the sensor is on the targeted sampling feature.)

Observation Process

An observation process is based on a method which is determined by the scientific definition of the property, the effects this property might have on other properties of the same feature-of-interest and the required accuracy of the obtained estimates.

The observation process may be a composite function (chain) consisting of an arbitrary number of functions (steps). As a consequence each process step is based on an underlying method and uses a ‘tool’ in a specific way.

A Process Step can either interact with (e.g. rain rate measurements) or refer to (e.g. prediction of rain rate) the feature-of-interest at elements of the property domain. Further, a Process Step can either be applied to provide a result (e.g. measurements or predictions of rain rate) or to create the appropriate conditions for a following Process Step (e.g. a specimen in an oven). Therefore, three main classes are proposed to represent all possible Process Steps from which any Process can be composed.

Acquisition: An Acquisition is a process step where one or more sensors (or human beings) interact with the feature-of-interest at elements of the property domain in order to provide property estimates.

Conditioning: A Conditioning is a process step which interacts with the feature-of-interest at elements of the property domain in order to create the appropriate conditions for a following Process Step. A Conditioning step can use instruments (or human beings) but does not acquire data.

Computation: A Computation is a process step where only numerical computations are involved, being executed by a machine or a human being, in order to provide property estimates at elements of the property domain. The computation does not interact with the feature-of-interest but refers to elements from the property domain.

The result of the observation should be escorted by data describing the observation process and its accuracy, the location and time that it took place. If the observation process is a composite function each of the individual Process steps should carry these properties.

Go  here for further details.

Observation Facility

An ‘Acquisition’ is always related to an instrument and an ‘Observation Facility’ which is a generic entity representing the facility where instruments are hosted or deployed to undertake environmental monitoring (in-situ or remotely sensed); either a static observation site or a moving platform. Examples of Monitoring Facilities are: stations, stationary platforms, aircrafts, satellites, High Elevated Platforms (HAPs), ships, tethered balloons etc. A Monitoring Facility can be a network consisting of one to any number of monitoring facility members.

A static observation site can host or operate zero to any number of stationary or moving platforms.

The location of a Monitoring Facility is always the location of the asscociated acquisition. Also there is no neeed of specialisation of different types of Acquisition as the type of Observation facility defines the type of the relevant acquisition . For example the acquisition which uses as an Observation Facility an aircraft is a flight, a ship is a cruise.

NOTE:The definition of Observation Facility can be expanded to represent also the facility where Conditioning and/or Computation Process Steps are performed.

Example: during the OP3 programme (project) measurements of ozone occurred ( observation). The observation procedure to obtain these estimates of ozone values (aquisition) used a Thermo Electron UV Analyser (instrument)at Bukit Atur station's (Observation Facility )100m tower (Platform).


The result of an observation is a function with domain the spatio-temporal extent of the feature-of-interest and range the corresponding property values (see (1) and (2)). The spatial variation of a result is modelled by the concept of coverage (ISO 19123) whereas for the temporal characteristics of a result ISO 19156 has a number of specialised observations including a specialisation for time series observations.

For discovery purposes, the result of an observation is defined uniquely from:

  • the pair (property , feature-of-interest)
  • the temporal sampling regime
  • the used procedure
  • the location of the procedure
  • the time at which the observation took place.

3. Observation Collection Section

The results can be appeared in collections which are organised with significantly more flexibility than would be done if one used the original project alone.

 Trac Powered
Site hosted at the
British Atmospheric Data Centre
for the