The paper presents estimation of positional accuracy of digital maps using statistical analysis. Investigations have been performed for four large-scale digital maps made using different methods of producing digital map data: new total station survey (object A), re-calculation of previous direct measurements (orthogonal and polar surveys) (object B), manual vectorisation of a raster orthophotomap image (object C) and graphical-and-digital processing of analogue maps (object D). Analysis has been performed for large statistical samples of sets of vectors of shift t: L of control points and their components, i.e. true errors t: x, t: r of increments of co-ordinates. In the case of a map produced by means of new survey with an electronic tacheometer, the true errors were represented by differences between co-ordinates of control points obtained from two separate set outs. In the case of other methods of data collection for digital map production true errors were represented by differences of co-ordinates acquired from an investigated map and co-ordinates calculated from new direct surveys.

Leak detection in transmission pipelines is important for safe operation of pipelines. The probability of leaks may be occurred at any time and location, therefore pipeline leak detection systems play a key role in minimization of the occurrence of leaks probability and their impacts. During the operation of the network there are various accidents or intentional actions that lead to leaks of gas pipelines. For each network failure, a quick reaction is needed before it causes more damage. Methods that are used to detect such network failures are three-staged-: early identification of leakage, an accurate indication of its location and determine the amount of lost fluid. Methods for leak detection can be divided into two main groups: external methods (hardware) and internal methods (software). External leak detection methods require additional, often expensive equipment mounted on the network, or use systems that could display only local damage on the pipeline. The alternative are the internal methods which use available network measurements and signalling gas leakage signal based on the mathematical models of the gas flow. In this paper, a new method of leak detection based on a mathematical model of gas flow in a transient state has been proposed.