NeuraCities
All posts
Urban ForestryStreet treesTree inventoryPlanting priorities

How to do a street tree inventory: what to collect, how to sample, and where to plant next

The NeuraCities Team 11 min read

A street tree inventory is the basic record of an urban forest program: what trees you have, where they are, how big they are and what condition they are in. Without one, pruning cycles, risk work, budgets and diversity goals are all based on guesswork.

Inventories can also cost more than they need to, and go stale fast, when nobody first decides what they are for. This guide covers how to scope one, what to collect, when a sample is enough, what aerial data can and cannot replace, and how to turn the record into a list of where to plant next.

What an inventory is for

Write down the decisions the inventory has to support before you choose a method. Different goals need very different data:

GoalWhat you needApproach
Annual work plans and pruning cyclesEach tree, with location and maintenance needComplete inventory
Managing riskEach tree of concern, assessed by a qualified personComplete or targeted inventory
A budget case or benefits estimateSpecies and size distributionSample may be enough
Checking species diversitySpecies for a representative populationSample or complete
Choosing where to plantExisting trees, open sites and needInventory plus GIS analysis

ISA's Best Management Practices for tree inventories make the same point: each attribute should relate directly to an objective. Every extra field adds time per tree, multiplied by every tree in the city.

Complete inventory or sample?

Complete inventorySample inventory
What it coversEvery tree in the defined areaAll trees on randomly selected street segments
AnswersWhich tree, where, and what to do about itHow many trees, which species, what size overall
Best forDay-to-day management and work ordersPlanning, budgets and first-time benefit estimates
Main drawbackHigher cost, and needs ongoing updatesCannot tell you about any individual tree

How big should a sample be?

Two published reference points are helpful. The i-Tree Streets sample inventory guidance says samples typically cover 3 to 6% of street segments, giving about a 10% standard error on the citywide tree count, depending on community size and how much segments vary. Virginia Tech's street tree assessment project randomly selects 5 to 15% of a locality's streets and counts all the trees on them.

Selecting segments at random matters more than the exact percentage. Choosing streets because they are easy to reach, or because they are known to be leafy, biases the result. Stratify by land use or neighborhood if your city varies a lot, since a single citywide estimate can hide exactly the gaps you care about.

A note on i-Tree Streets
i-Tree now lists Streets as a legacy tool that no longer receives science updates or compatibility support, and most of its features have moved to i-Tree Eco. i-Tree's FAQ names street segment sampling as the main feature not carried over. The sampling logic still holds, but check the current i-Tree tools before you commit to a workflow.

Define the population before you start

Decide what counts as a street tree: public right of way only, or also parks and other public land? What about private trees overhanging the street? Is there a minimum size? And should you record empty planting sites? Most of these choices cannot be changed later without going back out, and vacant sites are especially valuable, as the planting section explains.

What to collect for each tree

Massachusetts' tree inventory fact sheet sets the minimum as location, species, size and condition, with risk rating, pests, maintenance needs and site conditions as useful additions. Virginia's field protocol records species, trunk diameter, condition, age class, planting type and land use. A sensible starting schema looks like this:

FieldPriorityWhy it matters
Location (coordinates and street address)CoreCrews need to find the tree, and GIS needs to map it
SpeciesCoreDrives pest exposure, diversity and benefits
Trunk diameter (DBH)CoreA standard proxy for size, age class and benefit
ConditionCoreSets maintenance and removal priorities
Maintenance needCommonTurns an inventory into a work plan
Site type and widthCommonAffects what can be planted back
Overhead lines and hardscape conflictCommonImportant for species choice and sidewalk repairs
Risk ratingOptionalOnly useful when done by someone qualified, on a defined method
Pest or disease signsOptionalValuable for known threats such as ash in borer-affected areas
Vacant planting siteRecommendedA ready-made list for the planting program
Inspection date and inspectorCoreTells you how current each record is

If you plan to use benefit tools, check their input requirements before collecting. i-Tree Streets, for instance, required species, DBH and street segment ID for every tree. Species codes also matter: decide on one naming list up front, since fixing inconsistent species names later is tedious.

Risk ratings are not a side note

A condition rating is not a risk assessment. Under the ANSI A300 Part 9 and ISA framework, tree risk assessment has three levels: limited visual, basic and advanced. If your inventory records risk, state which level was used and who did it. Otherwise, a rating can imply more certainty than the survey delivered, and that can matter if a tree later fails.

Running the fieldwork

  1. 1Write a field guide with definitions and photos for each condition class, plus rules for edge cases such as multi-stem trees.
  2. 2Train and calibrate. Have crew members rate the same trees and compare their answers before they split up.
  3. 3Set up mobile collection with fixed pick lists, so species and condition are entered the same way each time.
  4. 4Start with a pilot of a few streets to test time per tree and find any problems with the schema.
  5. 5Audit a random subset by having a second person re-inspect some trees, especially for species and condition.
  6. 6Store the data in GIS or tree management software, with a clear owner and a defined update process.

Keeping it current

An inventory begins to age the day it is finished. The cheapest way to keep it current is to connect it to daily work: when a crew prunes, removes or plants a tree, the record updates. Virginia's assessment project, for example, accepted existing inventories only if they were relatively recent, within 5 to 10 years, and covered all streets. A staged cycle that reinspects a part of the city each year is also easier to fund than a single large resurvey.

Read the inventory for diversity risk

One of the first things to do with a finished inventory is check concentration. The usual benchmark is the 10-20-30 rule, attributed to Frank Santamour: no more than 10% of a tree population from one species, 20% from one genus and 30% from one family.

It is a floor, not a guarantee. Emerald ash borer shows why: a population within the genus limit could still lose a fifth of its trees. Santamour himself noted that the rule offers little protection against pests with a broad host range. Some programs set stricter limits, particularly for new planting. It also helps to check by neighborhood, since a citywide figure can hide a street where nearly every tree is the same species.

What aerial data and AI can and cannot do

Remote sensing and field inventories answer different questions. Knowing which is which keeps a project honest:

InformationFrom imagery or LiDARFrom the field
Canopy coverYes, and it is the best way to measure itImpractical across a city
Location of individual crownsYes, within the limits of the imageryYes, precisely
Public or privateYes, by overlaying parcels and right of wayYes, but slower
Height and crown sizeYes, with LiDARYes
SpeciesLimited, and needs validationYes
Trunk diameterNot measured directlyYes
Condition and structural defectsBroad signals at bestYes
Risk ratingNoYes, by a qualified assessor

Crown detection also has blind spots. Overlapping canopies can merge into one crown, and small young trees can be missed. So treat a detected tree count as a planning layer rather than a legal or work-order record, and verify it with field samples.

Remote sensing still earns its place: it gives a citywide picture in places that have no inventory, it helps target the fieldwork, and it can reveal gaps between neighborhoods before any crew goes out.

From inventory to where to plant next

An inventory says what you have. Planting priorities also need to know where trees can go and where they are needed most. That is a GIS question, and it takes three steps.

1. Find real planting sites

Vacant sites recorded in the field are the best source, because a person has seen them. Where you do not have those, you can derive candidates by removing existing trees, buildings, pavement, and utility conflicts, and spacing positions along the street. Field-check the shortlist either way, since buried services and other local conditions often differ from the data.

2. Measure need, block by block

A common pattern is to overlay canopy with heat and social vulnerability, so that planting reaches the people who have the least shade and the most exposure:

LayerTypical sourceWhy it matters
Canopy gapCanopy layer compared with a local targetShows where cover is lowest
HeatSummer land surface temperatureTargets hot neighborhoods. Note that surface and air temperatures differ
VulnerabilityCDC Social Vulnerability Index in the US, or a local equity layerDirects investment where people have the fewest resources to cope
Stormwater needImpervious share, combined sewer catchmentsLinks planting to drainage goals and budgets

American Forests' Tree Equity Score is a useful public reference. It scores neighborhoods from 0 to 100 by combining a canopy gap with a priority index of equally weighted climate, health and socioeconomic variables, and a lower score signals a higher priority. It is a good place to start, and a local analysis adds what a national score cannot: your actual planting sites, utilities and public land.

3. Publish how the ranking works

Weighting is a policy choice, not just a technical one. Decide the weights with the people affected, publish them, and show why each area made the list. An equity-led plan also has to budget for what happens after planting. Establishment watering and upkeep are part of the cost of a tree, and one that dies in year two delivers none of its promised benefits.

Using an inventory in NeuraCities Urban Forestry

NeuraCities Urban Forestry does not replace a field inventory. It works alongside one, and gives you a citywide starting point if you do not have one yet. Pick a place, upload what you have, and it finds the other layers.

  • Your inventory comes first. There is a typed upload slot for an existing tree inventory, and planting sites are carved around those trees. You can also upload utility lines, parcels and public land, and these take the place of public equivalents.
  • No inventory? Crowns are detected from imagery. A tree detection model locates crowns in 1 m aerial imagery, so new sites stay clear of them. This gives you a location layer, not a species, DBH or condition record.
  • Public and private are separated. Trees are split by ownership, and the analysis can keep to public land and street verges.
  • Candidate street tree positions are placed along each street, spaced by road class, from about 12 m on major roads down to 8 m on pedestrian streets. Sites are kept clear of buried pipes and existing trees.
  • Zones are ranked by heat, equity and canopy gap, and each has a stated primary driver, such as heat plus low canopy, equity priority, or stormwater permeability.
  • Each zone is costed: estimated trees, cooling, dollar value and grant-eligible share, with 20 years of costs and benefits.
neuracities.com
Urban Forestry analysis over Mount Pleasant, Vancouver: panels list the team's uploaded utility lines and tree survey, and seven layers the analysis found itself, including existing trees, heat, equity, parcels, public land, parks and streets.
An illustrative run over real Vancouver open data, from the Urban Forestry product page. The team supplies a tree survey and utility lines, and the analysis gathers the rest.
neuracities.com
Priority zones panel over Mount Pleasant, Vancouver: three ranked zones labelled heat plus low canopy, equity priority and stormwater permeability, with a zone card listing why it was chosen and what 26 new trees would give.
The same illustrative run: ranked zones, each labeled by its primary driver, with the reasons it made the list.

The product figures are 30 minutes from a boundary to a plan, and 25x cheaper than a traditional study. For validation, the canopy baseline measured 31.1% across Portland, Oregon, against a commonly cited 29.8%. These are model-derived, planning-level estimates, so confirm them with local field data. For the full canopy method, see how to build an urban tree canopy plan.

Every account's first analysis, up to 20 square miles, is free.

Frequently asked questions

What is a street tree inventory?

A record of the trees along a community's streets, usually including location, species, size and condition. It supports maintenance planning, risk management, budgeting, diversity goals and planting decisions.

What data should a tree inventory include?

At a minimum, location, species, size and condition. Common additions are maintenance needs, risk rating, pests, site conditions and vacant planting sites. Each field should relate directly to an objective, because every extra field adds time per tree.

What is the difference between a complete and a sample tree inventory?

A complete inventory records every tree and supports day-to-day management. A sample inventory records all trees on randomly selected street segments and estimates the whole population, which suits planning and budgets but cannot tell you about individual trees.

How many streets do you need for a sample tree inventory?

i-Tree's guidance says samples typically cover 3 to 6% of street segments, giving about a 10% standard error on the citywide tree count. Virginia Tech's assessment project samples 5 to 15% of streets. Random selection matters as much as the percentage.

Can AI or satellite imagery replace a tree inventory?

Not entirely. Imagery and LiDAR are excellent for measuring canopy and locating crowns across a whole city. Species, trunk diameter, condition and risk still need field inspection, so the two work best together.

How often should a tree inventory be updated?

There is no single rule. Keep records updated as crews prune, remove and plant, and consider reinspecting part of the city each year. Virginia's assessment project accepted existing inventories only if they were within about 5 to 10 years old.

How do you decide where to plant street trees first?

Combine real planting sites with measures of need, such as canopy gap, surface heat and social vulnerability, then rank areas and publish the weights. NeuraCities Urban Forestry ranks zones by heat, equity and canopy gap and shows the main reason each zone is on the list.

Can I upload my tree inventory to NeuraCities?

Yes. Urban Forestry has a typed upload slot for an existing tree inventory. Planting sites are then kept clear of those trees, and you can also add utility lines, parcels and public land.

More guides on urban forestry

Bring a boundary and whatever data you have, and get a canopy baseline and ranked planting zones.

Start with your trees

Have a question about your own data?

Talk to us, or read the common questions.