Where the type of a feature allows for a property that is dependent on some parameter, then the value of the property is a <i>function </i>of this parameter. <br /></p><p>EXAMPLE The length of a rail varies with temperature.<br /></p><p>If the variation is temporal or spatial, then the function is a <i>coverage</i> (CV_Coverage—ISO 19123) whose domain extent is the spatio-temporal extent of the feature. The value of a corresponding observation result shall therefore be a function or coverage, respectively. In practice, the observation will <i>sample</i> the relevant axis of the target feature, so the observation result is usually represented as a <i>discrete</i> function or coverage (CV_DiscreteCoverage).<br /></p><p>The target feature may have many observations made on it using different sampling regimes, so the sampling regime is associated with the act of observation, rather than being inherent in the feature of interest. This may be accommodated by the decomposition of the domain geometry (i.e. the CV_DomainObject elements) in the observation <i>result</i>. The decomposition of the domain geometry in the result provides an intrinsic element of the overall observation protocol<br /></p><p>NOTE The sampling regime may also be accommodated by multiple observations on a complex of sampling features (1.1.1.1).<br /></p><p>EXAMPLE The colour of a scene varies with position.The result of an observation of the property ”colour” of the scene is a coverage. Each domain element is a pixel whose index allows the spatial location within the scene to be obtained.<br /></p><p>EXAMPLE Many properties of an observation-well vary along its length, including rock-type, orientation, permeability etc.  These are conventionally encoded as “logs”, with different sampling regimes. Each well-log is a coverage whose domain is the curve describing the shape of the well. The domain is sampled with elements whose location is described in terms of 1-D position measured along the well axis. <br /></p><p>A simple case concerns sampling a property at points on an extensive feature. The observation result is a set of point-value pairs (CV_PointValuePair—ISO 19123).<br /></p><p>EXAMPLE Temperature may be sampled using an array of weather stations. The temperature field of the region covered by the array may be represented as a discrete point coverage, whose domain-elements correspond to the station locations. <br /></p><p>An important case concerns monitoring a time-varying property of a persistent feature by sampling at discrete points in time. The observation result is a set of time-value pairs (either CV_PointValuePair, in which the point geometry uses a temporal reference system, or CV_TimeInstantValuePair—OGC 06-188r1). <br /></p><p>EXAMPLE An air- or water-quality monitoring station observes properties such as ozone, turbidity, etc. The instantaneous value is a scalar concentration or index value. However, the value is time-dependent. The value may be expressed as a coverage whose domain is the period of interest.  This is usually described as a time series, which is a discrete time coverage.  <br /></p><p>The feature of interest may be naturally structured into elements, such as a road network composed of road-segments, or a state composed of administrative areas at a finer scale, or a farm composed of fields. Observation of a property of these features may capture its variation as a function of the sub-features. In these cases the standard members of the target feature are responsible for decomposition of the domain geometry. <br /></p>