Arson Science and Investigations

History

Arson inves­ti­ga­tion and evi­dence first entered the U.S. crim­i­nal legal sys­tem dur­ing the late 1800s. These ear­ly cas­es relied heav­i­ly on cir­cum­stan­tial evi­dence and eye­wit­ness tes­ti­mo­ny rather than sci­en­tif­ic analy­sis. The sci­en­tif­ic analy­sis of fire scenes began to gain accep­tance in courts dur­ing the 1940s, and by the 1960s, fire inves­ti­ga­tors were com­mon­ly admit­ted to tes­ti­fy as expert wit­ness­es, large­ly depen­dent on the individual’s qual­i­fi­ca­tions and expe­ri­ences. Arson inves­ti­ga­tors dur­ing this era relied large­ly on gen­er­al­ly accept­ed con­cepts to deter­mine whether a fire was acci­den­tal, includ­ing whether there was evi­dence of an accel­er­ant or the smell of gaso­line. Detractors argued that these con­cepts lacked sci­en­tif­ic rig­or. A 1977 gov­ern­ment report not­ed that com­mon arson indicators had received lit­tle or no sci­en­tif­ic test­ing” and that “[t]here appears to be no pub­lished mate­r­i­al in the sci­en­tif­ic lit­er­a­ture to sub­stan­ti­ate their validity.” 

A sig­nif­i­cant shift in the valid­i­ty of arson sci­ence occurred in 1992 with the pub­li­ca­tion of The National Fire Protection Association (NFPA) 921: Guide for Fire and Explosion Investigations. The NFPA 921 estab­lished the first com­pre­hen­sive, sci­ence-based guide­lines for fire inves­ti­ga­tion, and it became the nation­al stan­dard for fire inves­ti­ga­tion method­ol­o­gy. U.S. courts began to use NFPA 921 as a bench­mark to eval­u­ate expert tes­ti­mo­ny in arson cas­es, par­tic­u­lar­ly fol­low­ing the U.S. Supreme Court’s deci­sion in Daubert v. Merrell Dow Pharmaceuticals (1993), which estab­lished a strict stan­dard for admit­ting sci­en­tif­ic expert testimony. 

Several land­mark cas­es have now shaped how today’s courts con­sid­er arson evi­dence in crim­i­nal cas­es. The case of Cameron Todd Willingham, who was exe­cut­ed in Texas in 2004 despite seri­ous doubts about the cause of the fire for which he was con­vict­ed of set­ting, revealed that many of the sup­posed indi­ca­tors that point­ed to arson were based on out­dat­ed and invalid method­ol­o­gy. The case of Ernest Ray Willis, who was exon­er­at­ed less than a year after Mr. Willingham’s exe­cu­tion, sim­i­lar­ly demon­strat­ed how flawed arson­science could lead to wrong­ful con­vic­tions. Mr. Willis was exon­er­at­ed when mod­ern fire sci­ence revealed that there was no evi­dence of an intentional fire. 

The 2012 appel­late deci­sion in Han Tak Lee v. Glunt helped estab­lish prece­dent for chal­leng­ing the valid­i­ty of out­dat­ed fire inves­ti­ga­tion meth­ods and con­tributed to the even­tu­al exon­er­a­tion of Mr. Lee based on a more sophis­ti­cat­ed under­stand­ing of fire sci­ence. Mr. Lee had been con­vict­ed and sen­tenced to life in prison with­out the pos­si­bil­i­ty of parole for the arson mur­der of his daugh­ter. The United States Court of Appeals for the Third Circuit found that the prosecution’s fire expert tes­ti­mo­ny under­mined the fun­da­men­tal fair­ness of [Mr. Lee’s] tri­al because the tes­ti­mo­ny was premised on unre­li­able sci­ence and was there­fore itself unreliable.” 

Courts have increas­ing­ly required arson inves­ti­ga­tors to demon­strate adher­ence to sci­en­tif­ic meth­ods and to explain the sci­en­tif­ic basis for their con­clu­sions, not just their expe­ri­ence or visu­al obser­va­tions. Arson evi­dence pre­sent­ed in court today places greater empha­sis on lab­o­ra­to­ry analy­sis, doc­u­ment­ed and proven method­olo­gies, and con­sid­er­a­tion of alter­na­tive hypothe­ses. Many juris­dic­tions now rec­og­nize that advances in fire sci­ence may con­sti­tute new evi­dence that can sup­port appeals or habeas cor­pus peti­tions, par­tic­u­lar­ly in cas­es where con­vic­tions were based on now-dis­­­cred­it­ed arson inves­ti­ga­tion meth­ods. Challenges remain in ensur­ing that all juris­dic­tions con­sis­tent­ly apply cur­rent sci­en­tif­ic stan­dards to arson inves­ti­ga­tions, and the courts prop­er­ly eval­u­ate the reli­a­bil­i­ty of arson evidence.

Standard of Care 

The National Fire Protection Association’s 921 estab­lish­es a sci­en­tif­ic frame­work for fire inves­ti­ga­tions through sev­en sequen­tial steps: (1) iden­ti­fy­ing the prob­lem, (2) defin­ing the prob­lem, (3) col­lect­ing data, (4) ana­lyz­ing data, (5) devel­op­ing a hypoth­e­sis, (6) test­ing the hypoth­e­sis, and (7) select­ing a con­clu­sive hypoth­e­sis. But NFPA 921 remains a guide that inves­ti­ga­tors can choose to fol­low or dis­re­gard. As a result, the fire inves­ti­ga­tion pro­fes­sion lacks manda­to­ry stan­dards gov­ern­ing sci­en­tif­ic and valid methodologies. 

Flashover in a controlled environment

Pabloboj, CC BY-SA 4.0 <https://​cre​ativecom​mons​.org/​l​i​c​e​n​s​e​s​/​b​y​— s​a/​4.0>, via Wikimedia Commons

Fire pat­tern analy­sis serves as the pri­ma­ry method by which fire inves­ti­ga­tors deter­mine how burn pat­terns were cre­at­ed by exam­in­ing the shape, depth, tex­ture, loca­tion, and appear­ance of heat effects on walls, ceil­ings, floors, and fur­ni­ture. Using this analy­sis, in com­bi­na­tion with an under­stand­ing of fire behav­ior and dynam­ics, a fire inves­ti­ga­tor may deter­mine the area of the fire’s ori­gin and its cause. As fires burn longer, this process becomes increas­ing­ly more dif­fi­cult, as new pat­terns con­flict with and obscure ear­li­er pat­terns. One com­pli­ca­tion is the con­di­tion known as flashover,” which is the phase in a closed space where the tem­per­a­ture is so high that all com­bustible items begin to burn, even if far from the fire’s ori­gin. Quickly after flashover, the fire moves to full room involve­ment,” where the pro­gres­sion of the fire can cre­ate con­flict­ing burn pat­terns, mask­ing the ori­gin of the fire.

Arc map­ping is anoth­er tech­nique that fire inves­ti­ga­tors may use to sup­ple­ment oth­er meth­ods to deter­mine the spa­tial rela­tion­ship between points of dam­age and ener­gized elec­tri­cal con­duc­tors. However, NFPA 921 specif­i­cal­ly advis­es against using arc map­ping as a stand­alone method for deter­min­ing fire ori­gin, rec­om­mend­ing that it be used as a com­ple­ment to oth­er collected data.

NFPA 921 sup­ports the use of wit­ness infor­ma­tion as a valid source of infor­ma­tion in a fire inves­ti­ga­tion but offers con­tra­dic­to­ry infor­ma­tion regard­ing how inves­ti­ga­tors should use wit­ness state­ments in draw­ing con­clu­sions. While the impor­tance of wit­ness state­ments is acknowl­edged, the guide cau­tions that fire inves­ti­ga­tors must con­duct a thor­ough inves­ti­ga­tion and col­lect all pos­si­ble data to refute or sup­port witness assertions.

When search­ing a fire scene, inves­ti­ga­tors look for evi­dence of an accel­er­ant or ignitable liq­uid that may indi­cate the fire was inten­tion­al­ly set. Gas chromatography/​mass spec­trom­e­try (GC/​MS) pro­vides reli­able lab­o­ra­to­ry con­fir­ma­tion of ignitable liq­uids in fire debris sam­ples even with just trace amounts present. However, in a large fire scene, includ­ing a house fire, ignitable liq­uids are found in every­day, com­mon items and may unin­ten­tion­al­ly con­tribute to the burn pro­gres­sion of a fire.

Concerns with Arson Science and Investigations 

Fire inves­ti­ga­tors can deter­mine a fire’s ori­gin through pat­tern inter­pre­ta­tion when extin­guished dur­ing ear­ly stages. However, inter­pre­ta­tion becomes more com­plex as fires progress. Extended burn­ing cre­ates com­pet­ing lay­ers of evi­dence where new­er pat­terns obscure ear­li­er ones. This evolv­ing pat­tern com­plex­i­ty intro­duces sub­jec­tiv­i­ty into the inves­ti­ga­tion process, mak­ing accu­rate ori­gin deter­mi­na­tion increas­ing­ly chal­leng­ing as fire dam­age becomes more exten­sive. The lit­tle research that exists regard­ing area of ori­gin deter­mi­na­tion sug­gests com­plex fire scenes, when exam­ined by mul­ti­ple inves­ti­ga­tors, lead to different conclusions.

K9 Training with Florida State Fire Marshal’s Office

State Farm, CC BY 2.0 <https://​cre​ativecom​mons​.org/​l​i​c​e​n​s​e​s​/​b​y/2.0>, via Wikimedia Commons

In a 1986 pilot pro­gram devel­oped by the Connecticut State Police and the Bureau of Alcohol, Tobacco, and Firearms, police trained dogs to search fire scenes for traces of ignitable liq­uids. These dogs, who came to be known as accel­er­ant detec­tion canines (ADC) were trained to detect 17 dif­fer­ent odors of ignitable liq­uids. It is known that dogs’ noses are sen­si­tive enough to detect min­i­mal vol­umes, like those need­ed for GC/​MS, but it is unknown how well these ani­mals can dis­cern between actu­al ignitable liq­uids and oth­er chem­i­cal com­pounds that release sim­i­lar chem­i­cals after burn­ing in a fire. The ADCs often alert han­dlers to sim­i­lar-smelling com­pounds in burned mate­ri­als even in sit­u­a­tions where no actu­al ignitable liq­uids were used, reveal­ing a crit­i­cal lim­i­ta­tion in their scent discrimination capabilities. 

In a paper from Parisa Dehghani-Tafti, the Commonwealth’s Attorney for Arlington County and the city of Falls Church, and Paul Bieber, a crim­i­nal inves­ti­ga­tor and founder of The Arson Research Project, the authors assert that arson science is a sub­set of crim­i­nal inves­ti­ga­tions where the inves­ti­ga­tor con­sid­ers the total­i­ty of cir­cum­stances (includ­ing, per­haps, the con­clu­sions of foren­sic experts) to under­stand if a crime took place and, if it did, to gath­er evi­dence to prove each ele­ment, under­stand pos­si­ble motives, and iden­ti­fy a crim­i­nal sus­pect.” Because the sub­ject of a fire sci­ence expert’s tes­ti­mo­ny regards the ori­gin and behav­ior of a fire, these indi­vid­u­als also clas­si­fy the cause of the fire. This deter­mi­na­tion extends beyond where and how a fire start­ed but goes to the inten­tion­al or acci­den­tal nature of the fire. There are four clas­si­fi­ca­tions of fires: nat­ur­al, acci­den­tal, incen­di­ary, or undetermined. 

NFPA 921 com­bines foren­sic deter­mi­na­tions of fire ori­gins and cause with judg­ments of human intent when deter­min­ing the nature of a fire. This con­fla­tion allows inves­ti­ga­tors to present infer­ences about human behav­ior with the weight of seem­ing­ly sci­en­tif­ic tes­ti­mo­ny. Courts often per­mit these indi­vid­u­als to offer expert tes­ti­mo­ny that extends beyond tech­ni­cal analy­sis of a fire scene. NFPA 921 fails to sep­a­rate sci­en­tif­ic find­ings from inferred intent and cre­ates a prob­lem where the jury may con­flate this asser­tion with an objec­tive forensic conclusion. 

The National Academy of Sciences’ 2009 report, Strengthening Forensic Science in the United States: A Path Forward, rec­om­mends that foren­sic inves­ti­ga­tions should be inde­pen­dent of law enforce­ment efforts either to pros­e­cute crim­i­nal sus­pects or even to deter­mine whether a crim­i­nal act has indeed been com­mit­ted.” This sep­a­ra­tion is rec­om­mend­ed because stud­ies have demon­strat­ed that tak­ing on a par­tic­u­lar role can influ­ence an observer’s view­point, direct­ly affect­ing both the type of infor­ma­tion they seek and how they per­ceive and process that infor­ma­tion. This is rel­e­vant to fire inves­ti­ga­tors who may assume a crim­i­nal inves­tiga­tive role, as is the case in many jurisdictions. 

The NFPA 921 does not pro­vide strict guid­ance on how to sep­a­rate these roles. Rather, NFPA 921 lists arson indi­ca­tors to con­sid­er when assess­ing whether a fire was set inten­tion­al­ly and expands the phys­i­cal evi­dence which fire inves­ti­ga­tors should col­lect and con­sid­er. Some of these non-fire” indi­ca­tors include the absence of per­son­al items pri­or to a fire, evi­dence of addi­tion­al crime, and any indi­ca­tion of financial stress. 

Another issue that may unin­ten­tion­al­ly appear in fire inves­ti­ga­tion is expec­ta­tion bias — the ten­den­cy to favor, val­i­date, and empha­size results that align with antic­i­pat­ed exper­i­men­tal out­comes, while simul­ta­ne­ous­ly doubt­ing and reject­ing the impor­tance of find­ings that con­tra­dict these expec­ta­tions. This bias acts sim­i­lar­ly to con­fir­ma­tion bias, which is the ten­den­cy for an indi­vid­ual to seek infor­ma­tion that con­firms their pre­con­ceived notions. NFPA 921 acknowl­edges that both expec­ta­tion and con­fir­ma­tion bias may arise in fire inves­ti­ga­tions and warns inves­ti­ga­tors to be wary of pre­sump­tions, but the guide­lines do not offer tools on how to avoid this bias or rec­og­nize one’s own bias.

Case Studies 

Cameron Todd Willingham

Cameron Todd Willingham was con­vict­ed, sen­tenced to death, and ulti­mate­ly exe­cut­ed for inten­tion­al­ly set­ting a house fire that killed his three chil­dren. The pros­e­cu­tion’s case relied heav­i­ly on the tes­ti­mo­ny of two key fire inves­ti­ga­tors: Assistant Fire Chief Douglas Fogg of Corsicana, Texas and Deputy Fire Marshal Manuel Vasquez from the State Fire Marshal’s Office. Both inves­ti­ga­tors used what were then con­sid­ered stan­dard inves­tiga­tive tech­niques, iden­ti­fy­ing what they claimed were over 20 dis­tinct indi­ca­tors of arson, includ­ing burn pat­terns, melt­ed alu­minum they asso­ci­at­ed with accel­er­ant-fueled high tem­per­a­tures, and petro­le­um-based chem­i­cal traces found in sam­ples from the front porch floorboard. 

Mr. Fogg and Mr. Vasquez both relied heav­i­ly on visu­al inter­pre­ta­tion and what Mr. Vasquez termed log­i­cal and com­mon sense” obser­va­tions. Their process involved attempt­ing to elim­i­nate acci­den­tal or nat­ur­al caus­es of the fire while look­ing for what they con­sid­ered unusu­al burn char­ac­ter­is­tics and pour pat­terns. This approach, while stan­dard at the time, lacked rig­or­ous sci­en­tif­ic val­i­da­tion. Particularly prob­lem­at­ic was their use of com­par­a­tive analy­sis, with Mr. Vasquez liken­ing burn pat­terns to fin­ger­prints — sug­gest­ing they were unique and unchange­able iden­ti­fiers of arson, a claim that mod­ern fire sci­ence has thoroughly debunked.

Picture of the after­math of the Willingham’s house fire.

Photo from Texas State Fire Marshall’s Office.

Both inves­ti­ga­tors employed some sci­en­tif­ic test­ing, specif­i­cal­ly gas chro­­matog­ra­­phy-mass spec­trom­e­try (GC-MS), to ana­lyze sam­ples from areas of the house they deemed sus­pi­cious.” GC-MS test­ing failed tocon­clu­sive­ly demon­strate the pres­ence of accel­er­ants, but Mr. Vasquez argued that the absence of an accel­er­ant should­n’t be inter­pret­ed as evi­dence that accel­er­ants were not used — a posi­tion that effec­tive­ly dis­missed the sig­nif­i­cance of sci­en­tif­ic test­ing when it con­tra­dict­ed his conclusions. 

Mr. Vasquez was per­mit­ted to tes­ti­fy not only about the cause of the fire but also about Mr. Willingham’s intent and guilt. The inves­ti­ga­tors’ con­clu­sions were influ­enced by extrale­gal fac­tors, includ­ing wit­ness inter­views and Mr. Willingham’s per­ceived behav­ior after the fire, show­ing how inves­ti­ga­tor bias could impact the inter­pre­ta­tion of phys­i­cal evi­dence. Mr. Willingham was sen­tenced to death on October 29, 1992 and none of these issues were raised dur­ing appeals or in his ini­tial habeas corpus petition. 

Texas sched­uled an exe­cu­tion date for Mr. Willingham in 2004. Patricia Cox, Mr. Willingham’s cousin, enlist­ed Gerald Hurst, a fire and explo­sives expert, to review the evi­dence in his case. Mr. Hurst’s analy­sis, com­plet­ed as part of Mr. Willingham’s clemen­cy peti­tion, identified crit­i­cal errors” in the experts’ tri­al tes­ti­mo­ny and high­light­ed how sci­en­tif­ic advances chal­lenged each of the 20 arson indi­ca­tors that Mr. Fogg and Mr. Vasquez iden­ti­fied. By 2004, a new analy­sis indi­cat­ed that the fire that killed Mr. Willingham’s three chil­dren was acci­den­tal. Despite Mr. Hurst’s report detail­ing his find­ings, Mr. Willingham was denied clemen­cy and was exe­cut­ed by lethal injec­tion on February 172004

In 2008, the Texas Forensic Science Commission agreed to review the evi­dence in Mr. Willingham’s case. Fire expert Craig Beyler’s review for the com­mis­sion tes­ti­fied that the orig­i­nal inves­ti­ga­tors failed to con­sid­er sev­er­al plau­si­ble alter­na­tive expla­na­tions for the fire, includ­ing elec­tri­cal caus­es or acci­den­tal igni­tion by one of the chil­dren. Mr. Beyler con­clud­ed that nei­ther Mr. Fogg’s nor Mr. Vasquez’s inves­ti­ga­tions met the stan­dards of care that exist­ed even at the time of the orig­i­nal inves­ti­ga­tion, let alone the mod­ern stan­dards out­lined in the National Fire Protection Association’s guidelines. 

In 2009, The New Yorker released an in-depth arti­cle inves­ti­gat­ing Mr. Willingham’s case, draw­ing light on the many faults present in the ini­tial inves­ti­ga­tion. The arti­cle brought renewed atten­tion to Mr. Willingham’s case, and inspired the film, Trial by Fire, which details the events of the fire and time on death row ahead of his eventual execution.

Ernest Ray Willis

In October 2004, just eight months after Cameron Willingham’s exe­cu­tion, Ernest Ray Willis was freed from Texas’ death row large­ly because of changes in the under­stand­ings of fire science. 

Mr. Willis was con­vict­ed and sen­tenced to death for the 1986 arson mur­ders of two women. After a house fire killed Elizabeth Belue and Gail Alison, inves­ti­ga­tors quick­ly focused on Mr. Willis, who had been stay­ing at the house that burned down and was one of the only two sur­vivors. The pros­e­cu­tion’s case heav­i­ly relied on arson experts who claimed to identify pour pat­terns” on the floor, which they viewed as evi­dence of an inten­tion­al­ly set fire using flam­ma­ble liq­uids. Despite no clear motive or phys­i­cal evi­dence link­ing Mr. Willis to the arson, the pros­e­cu­tion por­trayed him as a cold-heart­ed” murderer and satan­ic demon.” 

In 2004, U.S. District Judge Royal Furgeson threw out Mr. Willis’ con­vic­tion, cit­ing sev­er­al issues includ­ing the fact that after mod­ern arson experts who reex­am­ined the evi­dence against Mr. Willis con­clud­ed that the orig­i­nal inves­ti­ga­tors had been fun­da­men­tal­ly wrong in their analy­sis. There was no evi­dence of inten­tion­al fire-set­t­ing, and the orig­i­nal experts had wrong­ly dis­count­ed the house’s doc­u­ment­ed elec­tri­cal prob­lems that could have caused an acci­den­tal fire. The cause of the fire was new­ly declared as unde­ter­mined,” high­light­ing how the orig­i­nal inves­ti­ga­tors jumped to con­clu­sions about arson with­out prop­er­ly con­sid­er­ing alter­na­tives. This analy­sis led pros­e­cu­tors to drop all charges against Mr. Willis, who was released from death row after sev­en­teen years in October 2004. He was even­tu­al­ly award­ed $429,000 in com­pen­sa­tion from the state, in addi­tion to a month­ly annu­ity of near­ly $10,000. Mr. Willis died in January 2021 at age 75.

Sources

Jennifer D. Oliva and Valena E. Beety, Evidence on Fire, North Carolina Law Review, 2019; Parisa Dehghani-Tafti and Paul Bieber, Folklore and Forensics: The Challenges of Arson Investigation and Innocence Claims, West Virginia Law Review, December 2016; Rachel Dioso-Villa, Scientific and Legal Developments in Fire and Arson Investigation Expertise in Texas v. Willingham, Minnesota Journal of Law, Sciences & Technology, 2013; Alexandra Gross, Ernest Ray Willis, National Registry of Exonerations, August 29, 2011; Innocence Staff, Cameron Todd Willingham’s Wrongful Execution Gains New Attention After Netflix’s Trial by Fire Release, Innocence Project, September 13, 2010 (Updated 4/​2/​2025); Jordan Smith, Innocent Man Off Death Row, The Austin Chronicle, October 15, 2004; John F. Boudreau, Quon Y. Kwan, William E. Faragher, Genevieve C. Denault, Arson and Arosn Investigation: Survey and Assessment, National Institute of Law Enforcement and Criminal and Law Enforcement Assistance Administration, 1977.