The land-sea mask (LSM) is an unchanging field containing the fraction of land within every grid box. The values lie between 0 (grid box is fully covered with water) and 1 (grid box is fully covered with land). A grid box is considered to be land if more than 50% of it is land, otherwise it's considered to be water (ocean or inland water, e.g. rivers, lakes, etc.). This binary choice of assignment of land/water points means that globally land is slightly under-represented in the model. The value of the land/water proportion strongly depends on the quality of used global land cover map and its horizontal resolution (current nominal resolution is ~300m). Difficulties also arise where small islands, parts of islands, or coasts cover half or less of a grid box. Smaller lakes may not be captured by the land-sea mask, and some lakes or inland seas vary in size and meteorological impact due to human or seasonal effects (e.g. the Sea of Azov where the depth varies substantially with the seasons). Some islands may be missed altogether (e.g. El Hierro in the Canary Islands). See also the discussion on the impact of the land-sea mask on the derivation of Meteograms which also uses the examples shown below.
Examples of grid point distribution and the effect on energy flux computation.
Grid boxes are coloured by the fraction of land cover - scales are on the right and apply to all figures in this sub-section.
Coastal Area - Isle of Wight, Southern England.
Fig2.1.7A: ENS grid points over part of southern England. Within each box all locations are considered to have the same values as forecast at the central grid point. The fluxes of heat, moisture and momentum at the grid point are calculated using the proportion of land (where the HTESSEL is used) and sea (shallow coastal sub-grid scale waters where FLake is used). Where the land fraction is <50% NEMO is used to provide oceanic fluxes unless the lake dataset specifically highlights the location as a lake (e.g. the Great Lakes) when FLake is used.
Consider the areas around the island shown in Fig2..1.7A (the Isle of Wight).
- Any location on and around the island within the green box will be represented by the single grid point near the centre. Within the grid box the land fraction is about 90% and the water fraction about 10%. Therefore HTESSEL will supply ~90% and FLake (rather than NEMO, because the land fraction >50%) supply ~10%.
- Locations on the island in the blue boxes will be represented by the corresponding grid points to the south or west. Within these grid boxes the land fraction is about 10%, and the water fraction about 90%. Therefore HTESSEL supply ~10% of the flux information and NEMO (rather than FLake, because the land fraction is <50% and the location is not a lake) will supply ~90%.
- Locations to the north of the island in the turquoise area will be represented by the grid points to the north and northeast. Within these grid boxes the land fraction is about 60% land and the water fraction about 40%. Therefore HTESSEL will supply ~60% of the flux information and FLake (rather than NEMO, because the land fraction >50%) will supply about 40%.
Fig2.1.7B: HRES grid points over part of southern England. More detail is represented than by the ENS grid, but note two points to the northeast of the island are considered as sea points (one of which encompasses the city of Portsmouth).
Fig2.1.8A: ENS grid points around Lake Geneva. Only one grid box has less than 50% land (blue) and any land locations within that box will be considered as if over water. FLake is used here because the lake dataset tells the IFS that this is a lake and not open ocean (i.e. different to the Southern England case shown above). The other turquoise shades show the proportion of land cover within the each box and define the proportional influence of the FLake and HTESSEL for any land point within the grid box. For example, a point on the northeast coast of the lake will use ~60% HTESSEL and ~40% FLake for evaluation of fluxes.
Fig2.1.8B: HRES grid points around Lake Geneva. Much more detail is captured by the grid and a more realistic representation of the influences of land and water within each grid box. NEMO is not used at all here because the waters are classed as a lake and not open ocean.
Fig2.1.9A: ENS grid points around the Canary Islands. Note several island points have a fairly high proportion of sea in their grid box, and some points on islands have no land within their grid box and are considered as sea points because their grid box has <50% land cover. The island of El Hierro (westernmost island on the plot) is not captured at all. Because this is an open ocean area in ENS all blue boxes on these plots NEMO is used to provide oceanic fluxes; all turquoise and dark green boxes use FLake and HTESSEL.
Fig2.1.9B: HRES grid points around the Canary Islands. Note there are many more points wholly on land. The island of El Hierro has one wholly land point and three partial land points with sea influence. Because this is an open ocean area in HRES all blue boxes on these plots use model sea-surface temperature data; all turquoise and dark green boxes use FLake and HTESSEL.