Half-mile grid cells (Uber H3 resolution 8: about 0.74 km² or 180 acres each, roughly half a mile across) whose center is in Broome, Chenango, Cortland, Delaware, Otsego, Schoharie, Tioga or Tompkins County: 21,111 in all. The layers are the ones on the
US grid map page, which maps each layer and describes its source.
Thresholds. Each numeric layer keeps the grid cells whose value lies between the two ends of its bar, set by dragging them. The defaults: employment access, activity, land use mix and intersection density start at the region's top 20% of grid cells (where many grid cells tie at zero, every grid cell above the tie, which can be fewer); price premium at an index of 1.00× or more, where the location adds to home prices (the Southern Tier map's home price index, drawn on these same grid cells); healthcare access at any hospital or pharmacy within half a mile of the center; transit frequency at 4 or more weekday 4-7 pm departures per hour from stops within a quarter mile, the Southern Tier map's frequent-transit test; flood zone at 10% or less of the land in FEMA's 100-year floodplain; and steep slopes at 10% or less of the land at 25% grade or steeper. The bars run on the same scales as the variable scores below, so the bar's ends past 0.6× and 2.0× take in every price premium beyond them. Water and sewer keeps grid cells with more than half their land in a public water service area, a public sewer service area, or both (the default), chosen with the switch in its row. FEMA's national layer has no digital flood map for most of Broome County (Binghamton included), so Broome grid cells read as outside the floodplain there. Grid cells that are 98% or more surface water (USGS NHD High Resolution lakes, ponds, reservoirs and wide rivers, wetlands left out) are hidden unless the box is unchecked.
Healthcare access adds, for each hospital or pharmacy within half a mile of a grid cell's center, 1 at the center falling to nothing at half a mile. Hospitals are National Structure Inventory hospital buildings (merged into campuses, 10,000 sq ft or more); pharmacies are OpenStreetMap and Overture places.
Variables. Each layer scores a grid cell from 0 to 100 from its own value, so layers on different scales count equally; the bars show that score and the figure beside them the value itself. Activity, intersection density, transit frequency and healthcare access run on a log scale from none (0) to the region's highest grid cell (100), employment access from the region's lowest grid cell to its highest, and price premium from 0.6× to 2.0× (the Southern Tier map's range, clamped beyond it; the score is used only to weight it, and the page always shows the index); the log scale spreads the small towns rather than crowding them near zero. Land use mix is uses present out of nine. Water and sewer scores the share of the grid cell's land with public service, water and sewer together counting in full and either alone counting half. Flood zone scores 100 less the share of the grid cell's land in FEMA's 100-year floodplain (the data lake's national layer, which has no 500-year zone), and steep slopes 100 less the share at 25% grade or steeper (the data lake's national USGS 3DEP layer, about 30 m, which tends to under-count small steep patches). A variable's range on its bar stretches this score: values at or below its lower end score 0, at or above its upper end 100, and those between spread evenly on the bar's scale. A grid cell's score is the average of the variables that are on, rounded to a whole number for its color: fixed classes of 0, 1-19, 20-39, 40-59, 60-79 and 80-100, lightest to darkest, so a color means the same score whichever variables are on. The 10-class blue splits the same 0 to 100 into ten classes of ten points (0-9, 10-19, up to 90-100) on the blue ramp stretched to ten colors, and the high-contrast colors into twenty classes of five points (0-4, 5-9, up to 95-100) on the design system's 20-stop ramp; the yellow-blue colors use the same twenty classes on CARTO's seven-color BluYl ramp, stretched to twenty.