Quick Projects · the trees of Washington DC
Washington has 10,405 flowering cherry trees that are nowhere near the Tidal Basin — on its streets, in its parks, outside its schools. 63% of them stand under a power line. That is not a coincidence, and it is the whole reason they are there.
Every spring the whole city goes to look at the same 3,800 trees. The Tidal Basin grove is the one everybody means by “the cherry blossoms,” it is about 70% Yoshino, and it fits in a ring you can walk in forty minutes.
The District's own tree inventory — every tree in public space, kept by the Urban Forestry Division — records 12,829 trees of the genus Prunus, of which 10,405 are ornamental flowering cherries. They are almost all somewhere else. Here they all are.
The inventory records, for every tree, whether there is an overhead utility line above it. For the city's 222,660 public trees as a whole the answer is yes 28% of the time. For the flowering cherries it is 63%.
Sorted by that share, the city's trees come out in a very orderly line, and the line is not about species — it is about height. Everything at the top is a small tree: lilac, Chinese pistache, cherry, crabapple, redbud, serviceberry, crape myrtle. Everything at the bottom is a tree that gets big: oak, elm, London plane, tulip poplar, sweetgum, pine.
This is a planting rule made visible. Where a wire runs over the verge, the forester cannot put in something that will grow into it, so they put in something that will not — a tree that tops out around twenty-five feet. In Washington that tree is usually a cherry.
Washington's eight wards do not share their wires evenly. Downtown and the old rowhouse core have almost none left overhead; the rest of the city has plenty. That split turns out to predict, almost exactly, where the cherry trees are.
In Wards 1, 2 and 6 — where barely one street tree in a hundred has a wire above it — cherries are about 2% of the trees. In Wards 3, 4, 7 and 8, where roughly four in ten do, cherries are 6% to 9%. Across the eight wards the two shares track each other at r = 0.92.
Which produces the joke this page is named for. Ward 2 contains the Tidal Basin, the Jefferson Memorial, the whole postcard. It also contains 270 flowering cherries of its own — the fewest of any ward in the city.
The Tidal Basin has one date. Peak bloom there is defined as the day 70% of the Yoshino flowers are open, it lasts several days, and the whole city plans around it.
The street trees do not work like that, because they are not all Yoshinos. Five cultivars make up almost all of them, and the published bloom order spreads them over about six weeks: Okame first, roughly a fortnight early; then the Higan cherries and the Snow Geese; then the Yoshinos, with the Tidal Basin; then, a clear two weeks behind everything else, the Kwanzans.
Two waves, with a real gap between them. On these windows the Yoshinos are over about five days after peak and the Kwanzans have not started for another three, so for the best part of a week Washington has almost no cherry blossom at all.
Drag the date. The map redraws with only the trees that should be in flower on that day, and the chart shows how many that is.
The inventory carries a planting date for 6,635 of the flowering cherries. Since 2004 the District has put in 288 a year on average, and the mix has kept changing. Snow Goose was a few dozen a year until about 2010 and then became a house standard, peaking at 192 in 2019. Okame barely registered through the late 2000s and has not been fewer than fifty in any full year since 2016.
The share going in under a wire has been falling the whole time: 94% of the cherries planted before 2005, 61% of those planted since 2020. Some of that is undergrounding, some is a wider planting list. The cherry is slowly becoming a tree Washington chooses rather than a tree it settles for.
The data. One source does nearly all the work: the District of Columbia's public-space tree inventory, layer 23 of the DC GIS Urban Tree Canopy service, maintained by the DDOT Urban Forestry Division. It is a live inventory of 222,660 trees, each with a species, a position, a trunk diameter, a condition, a ward, a lidar-derived crown, and — the field this page is built on — whether there is an overhead wire above it. Downloaded —.
What counts as a cherry. The species fields are free text and they are messy: the same tree is Prunus Okame, Prunus 'Okame' and Okame cherry in different rows, and a handful of records have a genus of Prunus with an oak or a crape myrtle written in the name. Names are normalised with a rule list you can read in src/classify.py; the mismatched rows are dropped. Of the 12,829 Prunus left, 10,405 are ornamental flowering cherries — the five named cultivars, the Japanese apricots, and 2,247 recorded only as Prunus, no cultivar. The rest are chokecherries, black cherries and purple-leaf plums, which are in the charts about wires but not in the ones about blossom.
What this page is not. It is not about the Tidal Basin. Those roughly 3,800 trees belong to the National Park Service and are not in the District's inventory at all, which is why the middle of the map is empty. Nothing here counts them, and the comparisons in the text are against the Park Service's own published figure.
The wire field. It is the inventory's own, recorded at survey time, with four values: none, low voltage, high voltage, both. This page counts the last three together as “under a wire.” The field is blank or unrecorded on a small minority of rows, which are counted as no wire — so the 63% is, if anything, a floor.
The thing that would be easy to get wrong. The first version of this analysis said cherries are planted under wires. That is true and it is not the finding, because every small tree is: lilac is higher than cherry, crabapple and redbud are just below it. The honest statement is the ladder in section 02 — the inventory is sorted by mature height, and the cherry's position in it is a consequence of being a twenty-five-foot tree, not of being a cherry. What makes it worth a page is the second step: because Washington's small-tree slot is filled with cherries specifically, the wire map and the blossom map are the same map.
Crown height. The inventory's lidar crown fields include a maximum height per crown polygon, and for cherries it runs to 127 feet — which is a neighbouring oak leaning into the polygon, not a cherry. Crown area is used here (it is what the 106 acres is) and crown height is not used at all.
Bloom dates. Sequence only, as the caption in section 04 says, and the whole season slides with whatever date the Tidal Basin peaks in a given year. Twenty-six records on a District-owned campus in Laurel, Maryland are dropped as not being in the city.