After the fire: which burned trees have to come down

No California rule sets criteria for assessing a fire-injured tree. PRC 4293 and GO 95 Rule 35 reach a burned tree only through the word dead, and CAL FIRE's field guide never discusses fire injury. The criteria in use come from a Forest Service marking guideline and a UC publication, imported by utility standards.

Three weeks after containment, a crew works a 12 kV circuit through the burn in a State Responsibility Area. One tree on the list is a ponderosa pine, 26 inches through at breast height, standing about 38 feet out from the line. It is upright.

The bark is blackened all the way around to roughly eight feet, and a little more than half the crown length is brown. Nobody on the crew argues about whether it is a problem. The record has to answer something narrower: whether it is a dead tree today, whether it is going to be one, and which of those two facts the rule is asking about.

That question is the whole of post-fire hazard tree work. The duties were written before the fire and do not change because of it. What changes is the number of candidate trees and the reliability of the word the duties turn on.

Two sentences already on the books reach a burned tree

The California stack has no post-fire provision. It reaches a fire-injured tree through two obligations that were there the day before the fire started.

Public Resources Code section 4293 closes with a duty that names four tree conditions, the first of which is "Dead trees," and requires that they be "felled, cut, or trimmed" to remove the hazard. That paragraph, and how far past the easement it reaches, is its own subject. GO 95 Rule 35 carries the Commission's version, attaching to trees that are "dead, rotten or diseased" and that overhang or lean toward a span, on a trigger that starts with the company's actual knowledge.

Both hinge on a word that a healthy forest makes easy and a burn scar makes hard. A green tree is not dead. A tree with no needles and loose bark is dead. The 26 inch ponderosa is neither, and it will not be either one for months.

Two consequences follow immediately. The first is that the burn scar itself is not knowledge of anything. Rule 35's clock starts on the date a company obtained actual knowledge of a specific condition on a specific tree, and a fire perimeter on a map does not supply that. The date somebody walked the span is the date the file carries.

The second is that a fire does not open a new category. A tree either meets the words in the two provisions or it does not. Where a hazard tree finding does have a defined consequence, in 14 CCR 1257(b), the finding withdraws an exemption rather than creating a duty.

The stack says dead and never says how to tell

Search the 2021 California Power Line Fire Prevention Field Guide for the vocabulary of fire injury and it returns nothing. The word "scorch" does not appear. Neither does "mortality," "salvage" or "fire-injured." The word "cambium" appears twice, once in a list of root disease indicators and once in a passage about dwarf mistletoe swellings, and neither has anything to do with heat.

That is the document the state stack points to for tree condition. Its defect catalog is built for a forest that has not burned: heart rot, butt rot wounds at the ground line, dwarf mistletoe and rust cankers, dead tops, human interference with growth. What it does and does not set out on hazard trees generally is that section's own ground, and on a charred bole it is silent.

Article 4 of Title 14 does not fill the gap either. Its twelve sections define eight terms and none of them is a fire injury term.

So the criteria are not in the rules. They are in documents the utility chooses, and the one thing the record can be held to is which document was used and by whom.

Where the criteria actually come from

PG&E publishes both halves of its post-wildfire program, and they are the clearest public answer in California to a question no California rule answers.

Utility Standard TD-7114S, Revision 1, effective July 15, 2024, records where it came from in its first requirement:

As part of its 2022 Wildfire Mitigation Plan (WMP), the Office of Energy Infrastructure Safety (OEIS) required PG&E to develop a standard to be used in a post-wildfire response.

Its Appendix A, "Vegetation Assessment Within the Wildfire Footprint," is one page long and sets the floor:

The assessment must include a minimum of the following criteria: - Species identification (genus or common name) - Crown condition assessment - Visual cambium condition assessment

Appendix A then names four documents, and three of the four were written for somebody else:

• Smith and Cluck's 2011 marking guidelines for fire-injured trees in California

• the University of California's Burned Oaks: Which Ones Will Survive?

• a Montana State University Extension photo-guide for the Northern Rockies

• a Forest Service post-fire assessment for conifers in Oregon and Washington

Utility Procedure TD-7114P-01, Revision 0, effective August 6, 2024, turns that into a field decision with three priorities and a job aid, TD-7114P-01_JA01, the "Wildfire Hazard Tree Evaluation Job Aid." The crown measure it uses is the Percent Crown Volume Scorch Ratio, PCVR.

Priority	Crown and char	Expected mitigation
F1 Immediate	PCVR of 65% or greater and/or Heavy-Severe bark charring	within days or weeks of identification
F2 Urgent	PCVR of 65% or greater and/or Heavy-Severe to Moderate-Significant bark charring	within weeks or months
F3 Non-Urgent	less than 65% PCVR and/or Light-Minor bark charring	within months

The F codes are the company's own triage. The Rule 18 priority levels that carry the Commission's correction periods run on a separate scale, so a post-fire file reporting F1, F2 and F3 answers a different question from the one an auditor asking about a Level 1 has in mind.

The qualifying conditions under F1 are worth reading as a list, because four of the five are structural rather than physiological: hollows, vertical cracks, uprooting, burnt-out roots, and blocking access of restoration personnel. A fire changes the base of a tree, and the base is where the assessment has to go.

Crown and cambium, and the month you read them

Marking Guidelines for Fire-Injured Trees in California, report RO-11-01 of May 2011, by Sheri L. Smith and Daniel R. Cluck of Forest Service Region 5, is the document underneath most conifer assessment in this state. Its first page says what it was written for, and the sentence is worth quoting before anything is built on it:

If you are using these marking guidelines for your post-fire restoration it is imperative that you contact your local Forest Health Protection (FHP) service area staff for review of your draft NEPA document (before public distribution), responses to comments and for assistance with marking guideline selection and project implementation.

That is a federal land management document, keyed to a federal environmental process. No CPUC general order and no CAL FIRE regulation adopts it. It is in California utility practice because a utility standard put it there.

The guidelines measure crown injury as a percentage of pre-fire crown length, and the variable changes with the season. Before bud break the readable number is percent crown length scorched, because, in the report's words, "heat killing of foliage may occur with only light injury to buds and twigs and the full extent of crown kill cannot be determined until bud break occurs."

After bud break the readable number is percent crown length killed. They are different measurements of the same tree and they carry different thresholds. For yellow pine of 10 to under 30 inches diameter at breast height, at a 0.70 probability of mortality, the post-bud-break threshold is 60 percent crown length killed and the pre-bud-break threshold is 90 percent crown length scorched.

Cambium is sampled, not estimated. The method is four hatchet cuts:

Sample cambium in as small an area as possible at 4 equally spaced locations around bole and within 3" of the ground line to minimize tree wounding

Dead cambium is "darker in color, often resin soaked and hard or gummy in texture." Live cambium is "lighter in color, moist and rather pliable." The count of dead samples, zero through four, is the cambium kill rating, and bark char class can substitute for the cut on unburned, lightly charred and deeply charred quadrants.

Run the 26 inch ponderosa through it. Assessed after bud break on crown injury alone, at 55 percent crown length killed, it sits below the 60 percent threshold and stays. Add the cambium rating and the threshold moves: add five points at a rating of 0 or 1, no change at 2, subtract ten points at 3 or 4.

At a rating of 0 the threshold becomes 65 and the tree stays. At a rating of 4 it becomes 50 and the tree is marked. Four cuts with a hatchet, three inches off the ground, decide a 26 inch pine that looks identical either way.

Oaks are answering a different question

Much of the California wildland urban interface corridor is oak, and oak does not behave like pine under the same fire.

UC ANR Publication 8445, Burned Oaks: Which Ones Will Survive?, by Doug McCreary and Glenn Nader, January 2011, covers blue oak, interior live oak and black oak. Its central finding inverts the conifer arithmetic:

However, even if a small portion of the circumference of the cambium remains alive (as little as 10%), the tree will likely survive, though the overall vigor could be reduced considerably.

Crown injury carries much less weight here. A tree that lost all its foliage will likely leaf out next spring if the cambium held. The diagnostic moves to the bark: cracked or separated bark means the cambium is "almost certainly dead," dark or yellowish cambium is probably dead, white or pink is most likely alive.

The publication's cut list is keyed to diameter and basal char rather than to crown:

• under 6 inches in diameter and scorched all the way around the base

• 6 to 12 inches with continuous charring around the base and reductions in bark thickness

• over 12 inches with continuous charring, pronounced reductions in bark thickness and occasional exposure of underlying wood

• basal wounds on 50 percent or more of the trunk, in residential or recreational areas presenting a mechanical risk

Read the last one closely. It is a target test, and the targets it names are people and structures. Conductors are not in it, because the publication was not written for a utility corridor. Whoever imports this list into a line clearance program is supplying the target themselves.

One more oak behavior belongs in the post-fire prescription rather than the assessment. A killed oak that is cut will sprout, "especially vigorous for live oaks," and what comes back stands in the same place under the same span. A sprout carries its own growth rate, and a removal that ends at the stump has produced next cycle's clearance problem in the same coordinates.

The salvage mark is set to a management objective

This is the place where a post-fire program most often goes wrong, so it is worth stating plainly. The marking guidelines do not identify hazardous trees. They sort trees by probability of mortality against an economic and silvicultural objective, and the threshold is chosen by the land manager:

Use lower probability thresholds when it is important to prevent leaving trees that may die Use higher probability thresholds when it is important to prevent taking trees that may live

A probability of mortality of 0.7 is a management setting, picked for a stand by the person running the project, and the report says as much.

The report says so itself, in a footnote attached to both fir tables:

FHP monitoring of fire-injured white fir revealed high levels of decay developing where significant cambium kill occurred at the root collar and on the bole. A portion of these decayed trees failed during the five-year period while still retaining green foliage. Land managers should be aware that even though true firs with high levels of cambium kill have a high probability of survival they may become hazards to people or property.

A white fir with a cambium kill rating of 4 and a green crown comes out of the mortality model as a survivor, and out of the failure question as the tree most likely to be lying across the conductor in year three. A program that runs the mortality tables and stops has filed the wrong finding for the right tree.

Three documents, three clocks, and none of them is a regulatory deadline:

Document	What it times	The interval
Forest Service RO-11-01	evaluation of a fire-injured conifer	before the beginning of the second post-fire winter, preferably within the first post-fire year
UC ANR 8445	reading cambium injury	several weeks after the fire
UC ANR 8445	final decision on a large oak of uncertain crown survival	at least one and preferably three growing seasons
PG&E TD-7114P-01	mitigation of an F1 Immediate tree	days or weeks of identification

The utility's clock is the fastest of the four, and the biology's is the slowest. Nothing in the Public Resources Code, GO 95 or Title 14 reconciles them. The utility picks, and the defensible version of that pick is a dated finding that says which document produced it and when the tree gets looked at again.

Felling the tree and selling the log are different permits

A crew can fell a fire-killed tree under the easement or under the access the Public Resources Code grants. What happens to the wood is a separate regulatory question, and it has a clean answer.

PRC 4527 defines timber operations as "the cutting or removal, or both, of timber or other solid wood forest products, including Christmas trees, from timberlands for commercial purposes." The same section defines commercial purposes to include trees processed into logs, lumber or other wood products offered for sale, barter, exchange or trade.

Two limits sit inside that first sentence. The land has to be timberland under PRC 4526, which is land available for and capable of growing a commercial crop of trees. And the purpose has to be commercial. A hazard tree felled and left, lopped or chipped is outside the definition entirely.

Sell the logs off timberland and the Forest Practice Act applies. The utility's own route through it is 14 CCR 1104.1(c), which covers clearing by a private or public utility "for construction of gas, water, sewer, oil, electric, and communications... rights-of-way, and for maintenance and repair of the utility and right-of-way." It is filed on CAL FIRE form RM-73 (1104.1(b)(c)), "Public Agency, Public or Private Utility Right of Way Exemption," revised April 2025, through the CalTREES system. Four facts on that form set the pace of the work:

• The Director has fifteen days from receipt to determine whether the notice is complete and accurate.

• No timber operations may be conducted until CAL FIRE's notice of acceptance is received, and a copy of the notice and the acceptance stays on site during operations.

• The notice is valid for one year from the date CAL FIRE accepts it, and operations must be complete within that year.

• The form asks for the Licensed Timber Operator by name and license number, the timberland owner of record, and the timber owner of record.

The width cap is where this collides with the felling duty. Section 1104.1 limits the exemption to its Table of Normal Rights-of-Way Widths, which gives 20 feet for electric distribution and transmission single circuits on poles at 0 to 33 kV and 45 feet at 34 to 100 kV. The PRC 4293 obligation is set by height and lean and reaches past the easement.

So a fire-killed sugar pine 45 feet off a 12 kV line is squarely inside the felling duty and squarely outside the 20 foot width the exemption covers. Felling it and leaving it raises nothing. Hauling its logs to a mill needs another door, and 14 CCR 1038(b) is the usual one, exempting dead, dying or diseased trees "in amounts less than ten (10) percent of the average volume per acre."

PG&E's standard puts the same instruction to its own crews at 2.8:

The scope of work should consider conformance with applicable Forest Practice Rules and potential tree protection ordinances. Exception is when the Governor has declared an emergency exemption to certain activities.

Where the wood is left, one detail crosses back into the clearance stack. Slash stacked against a subject pole is flammable material inside the ten foot firebreak PRC 4292 requires, and a burned-over site that satisfied the firebreak on the day of the fire stops satisfying it the moment the felled tree lands in it.

The line stays off until the F1 work is done

The one genuine emergency in post-fire vegetation work is a tree that can reach a conductor the restoration crew is about to energize. PG&E's procedure states the outcome directly:

F1 Immediate priority trees are prioritized for tree work completion and should be mitigated before Maintenance & Construction (M&C) energizes the overhead line.

The verbs in the operating steps are the honest part. The vegetation lead is told to REQUEST that affected lines remain de-energized while F1 work is pending, and, if necessary, to ASSERT to the Incident Commander that the work has to be complete first. The vegetation side does not own the switch, and no California rule gives it to them. The decision sits with the incident command structure, which is why the request and the response to it belong in the file rather than in somebody's memory.

Access is the other thing a fire changes. After an evacuation the landowner may be unreachable for weeks, and a refusal and an unobtainable permission are two different files. PG&E's standard says notification "should" follow normal inspection procedures and PRC 4295.5 "when possible and practical," which is an honest description of the conditions and not a substitute for the notice the statute requires when the statute applies.

Nothing in the stack sends you back down the burned circuit

The tree that stays is the tree the program has to catch later, and no interval in the California stack is attached to it.

GO 165 contains no vegetation row, and GO 95 Rule 31.2 routes supply line inspection frequency to GO 165. So the burned span assessed in October is next due whenever the utility's vegetation cycle says, and that cycle comes from the wildfire mitigation plan rather than from a general order.

The plan does have to speak to this. PG&E's 2026-2028 base wildfire mitigation plan carries a section 9.10, "Post-Fire Service Restoration," and reproduces Energy Safety's instruction above its own answer:

In this section, the electrical corporation must describe how post-fire service restoration vegetation management are scheduled (or triggered). This must include how the schedule is affected by HFTD tier or other risk designation.

The answer filed against that prompt describes an Initial Phase and an Extended Phase. It contains no interval, no trigger date and no tier-dependent number. That is the state of the art in California as filed, and it means the reassessment date on a fire-injured tree is set by the utility and recorded, or it is not set at all.

The procedure does name one escalation. Larger diameter F3 Non-Urgent trees "should be reassessed by an International Society of Arboriculture (ISA)-certified arborist, a Tree Risk Assessment Qualified (TRAQ) arborist, or Registered Professional Forester (RPF) to confirm probability of mortality." That is a second look by a named credential, on trees the first look decided to keep, which is exactly the population the fir footnote warns about.

What the record has to name

There is no post-fire tree form in the California stack. What follows is what a finding on a burned tree has to carry to still mean something when it is read against a records request two years later, and every field is the program's own choice.

The assessment date, and the containment date beside it. A crown reading has a shelf life measured in months and a season attached to it. "October 14" is a fact. "October 14, 22 days after containment, pre-bud-break" is a finding somebody can evaluate.

The species at species level, and the measure used. Percent crown length killed, percent crown length scorched and percent crown volume killed are three different numbers, and the species decides which one applies. A record that says "60 percent scorched" without saying which measure, on a tree identified only as "pine," cannot be re-checked.

The cambium evidence. The number of dead quadrants out of four, or the bark char class that stood in for the cut. PG&E's procedure requires the PCVR and the killed quadrant count to be entered "into the Comments text field in the system of record," which is a free text box carrying the two numbers the whole finding rests on. Any program that can put them in their own fields should.

The document and the threshold. Which guideline produced the call, and at what probability of mortality level. A mark made at 0.7 and a mark made at 0.5 are different decisions about the same tree, and the guidelines say to document the rationale for the level chosen.

The target, in the utility's own terms. Height, distance to the nearest phase, and lean. None of the imported documents supplies this, because none of them was written about conductors. The reach question decides the finding before any defect does, and it is the part of a post-fire file most often missing.

The reassessment date for every tree left standing. Nothing in GO 95, GO 165, the Public Resources Code or Title 14 sets one. A tree assessed as a survivor in the first post-fire year, with a cambium kill rating of 3 or 4, is a scheduled question rather than a closed one. If the file does not carry the date, the next person to stand under it is reading a finding with no expiration on it.

Further reading

• Hazard trees: what CAL FIRE's field guide actually says: The reach and failure inquiry a post-fire finding still has to satisfy, and the edition question behind it.

• PRC 4293: radial clearance in State Responsibility Areas: The felling duty a fire-killed tree lands in, and the four tree conditions the statute names.

• Taking a tree down immediately: the authority and its limits: What a Level 1 finding permits on somebody else's land, and where an emergency proclamation stops.

Frequently asked questions

Does California law say how to decide whether a fire-damaged tree will die?

No. Public Resources Code 4293 and GO 95 Rule 35 reach a burned tree through the word dead, and neither supplies a criterion for reading a scorched crown or a charred bole. The 2021 California Power Line Fire Prevention Field Guide, which is the only document in the state stack that treats hazard trees at length, contains no occurrence of the words scorch, mortality or fire-injured. The criteria in use are imported from a US Forest Service marking guideline and a University of California publication by the utility's own standard.

How long after a fire should a burned tree be assessed?

The three documents in use give three different answers. The Forest Service marking guidelines say all trees should be evaluated before the beginning of the second post-fire winter, preferably within the first post-fire year. UC ANR Publication 8445 says cambium injury should be read several weeks after the fire, and that a final decision on a large oak whose crown survival is uncertain is better left one to three growing seasons. PG&E's post-wildfire procedure expects its F1 Immediate trees to be mitigated within days or weeks of identification.

Is a fire-killed tree in the right of way a timber operation?

Only if it is removed for commercial purposes. PRC 4527 defines timber operations as cutting or removal from timberlands for commercial purposes, and defines commercial purposes to include logs processed into wood products offered for sale. A hazard tree felled and left, lopped or chipped is outside that definition. The same tree bucked into saw logs and sold is inside it, and on timberland it needs the utility right of way exemption on CAL FIRE form RM-73 or another exemption.

Does the utility right of way exemption cover a hazard tree outside the corridor?

No. 14 CCR 1104.1(c) limits utility clearing to the widths in the Table of Normal Rights-of-Way Widths, which is 20 feet for single circuits on poles at 0 to 33 kV. The PRC 4293 felling duty is set by tree height and lean and can reach beyond the easement, so a fire-killed tree tall enough to strike the line from 45 feet is inside the duty and outside that width. Removing its wood commercially rests on a different exemption or a plan.

Is there a rule requiring a utility to re-inspect vegetation on a burned-over circuit?

Not by interval. GO 165 carries no vegetation row, and GO 95 Rule 31.2 routes supply line inspection frequency to GO 165. Energy Safety's wildfire mitigation plan guidelines do require a Post-Fire Service Restoration section describing how post-fire vegetation work is scheduled or triggered, and PG&E's 2026-2028 base plan answers that prompt with two phases and no date. The reassessment date is the utility's own number.