CEH Integrated Hydrological Digital Terrain Model (IHDTM) comprises five 50 m grids: elevation, surface type, outflow direction, inflow directions and cumulative catchment area. It covers the UK and the areas of the Republic of Ireland that drain into Northern Ireland. It has been derived, using CEH algorithms, from the following OS digital 1:50,000 map data: contours, heighted lake shores, high water line and selected hill-top spot heights (marketed as PANORAMA vectors by OS GB), together with CEH’s digital river network of Great Britain.
The principal stages in the production of IHDTM were: 1) Heighting the digital rivers by non-linear interpolation between their intersections with contours, lake shores and coastline. Due to positional inconsistencies between some rivers and contours, a small proportion of the rivers have not been heighted. 2) At every point on a 50 m square grid, interpolating between the heighted rivers and the above-listed OS datasets (treating the high water line as a 3 m contour) to obtain an elevation grid that is spatially consistent with the river network. The method used takes a weighted average of elevation estimates obtained from multiple transects which intersect at the point, with the weight for each estimate being based on the confidence in the interpolation. 3) A by-product from the elevation gridding algorithm is the surface type grid. The value at each point is the most significant hydrological type in the 50 m x 50 m square centred on the point, from – in increasing order of significance – land, river, lake and sea. Thus a grid point may lie on land, but be classified as sea or lake in order to ensure the continuity of narrow (sub grid interval) estuaries and lakes within their gridded representation. Rivers in this grid are sometimes missing because the method uses heighted rivers and, as noted in (1) above not all rivers have been heighted. For a more comprehensive grid representation of the rivers, the outflow grid should be used (see (4) below). 4) The outflow grid was derived by (a) representing the digital rivers as a continuous sequence of vectors on a 50m square grid; (b) in each lake, setting the directions at all the contained points so that they lead to the lake outlet; (c) setting the direction at all the other land points by reference to their elevation and the elevations of their eight near-neighbours; the algorithm used was designed to give a sequence of flow directions that follow the true plan direction of steepest slope and therefore did not set every individual point to flow to the neighbour that gave the steepest descent, as this would have led to sequences of points artificially following one of the eight cardinal directions; finally, where necessary, flow was directed uphill to provide continuity from local low points. 5) The inflow directions grid was derived from the outflow grid, and for each point it lists which of its eight near-neighbours flow towards it. The main purpose of this grid is to enable rapid derivation of catchment boundaries using a CEH algorithm. 6) The cumulative catchment area grid was derived from the outflow grid, by recording how many points there are at or upstream of each point