Covariance indicates the relation between two variables with different units of measurement. A positive covariance means the two variables move in the same direction and have a positive relationship while negative covariance means that the variables move in opposite directions and have an inverse relation.
However, covariance cannot measure the degree to which the variables move together because the unit of measurement is not the same.
Trenchlesspedia Explains Covariance
Covariance and correlation are two different ways to determine the relation between two variables. While covariance cannot give a measure of the degree to which variables move together, correlation standardizes the measure of interdependence between two variables and tells how closely the variables move.
To measure correlation, it is necessary to know the covariance between the two variables and the standard deviation of each.
Correlogram, also known as Auto Correlation Function (ACF) plot is a graphic way to demonstrate serial correlation in data that does not remain constant with time. Correlograms gives a fair idea of auto-correlation between data pairs at different time periods.
It is used as a tool to check randomness in a data set which is done by computing auto-correlations for data values at different time lags. The auto-correlations are near zero for any time lag separation if it is random but if not, one or more of the auto-correlations will be non-zero.
Covariance and correlation are used to analyze market returns and to understand the interdependence between consumer behavior and consumption of the product. Correlograms can be used by municipalities to make estimates regarding repair or replacement or up-gradation of water and wastewater pipelines.
Correlograms are also used for forecasting sewage inflow into wastewater treatment plants and to determine annual variation in properties of concentrated sludge.