How UCLA Became the Only UC Campus to Cut Its Dedicated GIS Librarian
Author
Affiliation
Tim Dennis
UCLA Library Data Science Center
Published
March 24, 2026
Modified
September 19, 2026
NoteUpdate
In August 2026, UCLA committed to hiring a dedicated GIS librarian.
ImportantFinding and Recommendation
In 2024, UCLA cut its dedicated GIS librarian position. It’s the only UC campus that previously had a dedicated GIS librarian and eliminated the role, and the data in this report show it’s an outlier among national research peers as well.
The fallout is documented. By December 2025, UCLA’s ArcGIS Online platform had grown to 230–240% of its licensed storage capacity with no one managing it: no governance, no retention policy, no designated owner. An internal OARC report to UCLA leadership listed the vacant GIS librarian as a root cause. The proposed patch, buying more storage from Esri, runs $32,400/year and doesn’t restore any of the support the position actually provided.
Recommendation: Restore the GIS librarian position. At a UC salary of roughly $80,000–$115,000/year, restoring the position costs roughly twice UCLA’s current annual spend to maintain the platform in its degraded state ($47,400/year for site license plus storage patch), and it actually resolves the problem rather than deferring it.
WarningWhat is breaking down now
This isn’t a future risk. The gaps are active:
ArcGIS Online reached 230–240% of licensed storage capacity. UC-system negotiations with Esri on the license are underway but unresolved. OARC has taken on some platform responsibilities, but without a designated owner the support layer - consultations, instruction, data management, domain expertise - remains absent. Lowering the barrier to entry is not the same as provisioning the support.
LARIAC access: the Data Science Center still provisions EagleView accounts, but no one can help researchers understand what they’re looking at. Domain expertise left with the position: which LiDAR vintage to use, how to process point clouds, what each product is good for.
Research consultations: averaged 72-82 per year across urban planning, ecology, public health, digital humanities, and community organizations through the final year of the position. No equivalent GIS consultation service has been established since the position was vacated in August 2024.
GIS instruction: a for-credit GIS course co-taught with the Fielding School of Public Health, guest lectures across 4+ departments, library workshops reaching 400+ attendees, and UC-wide workshop organizing (617 attendees in the 2023 cycle alone). We are not aware of any unit that has announced equivalent programming since.
Data lifecycle support: metadata, provenance documentation, and preservation planning for spatial datasets produced by UCLA researchers. Without a librarian managing it, the datasets are still being created; they’re just no longer being described in ways that make them findable or reusable.
Overview
This report documents the presence or absence of dedicated GIS librarians at all 187 Carnegie R1 research institutions in the United States. Data were collected through systematic review of institutional library staff directories, subject guides, and library websites in March 2026.
82
R1 institutions with a dedicated GIS librarian
110
individual GIS librarians and specialists identified
81%
of the top 25 R&D-spending universities have dedicated GIS support
105
R1 institutions with no dedicated GIS librarian
Note
A note on special-focus institutions: Twenty of the 187 R1s are medical schools, health science campuses, or single-discipline institutes like Caltech, NJIT, and RPI. Health sciences campuses in particular represent a structurally different model, where GIS tends to sit outside the library entirely when it exists at all, so their absence here isn’t directly comparable to a comprehensive R1’s staffing decisions. Including them pulls the overall average down in a way that obscures what’s normal for a comprehensive research university.
The comparisons that matter for UCLA are with the institutions it actually competes with. Among the top 25 R&D spenders nationally, nearly all have dedicated GIS staff. Among UC comprehensive campuses, 6 of 9 do. Among UCLA’s direct peers in the US News public rankings, every single one does. Each is documented below.
What counts as a “dedicated GIS librarian”?
For this analysis, a position qualifies if GIS, geospatial data, maps, or spatial data appear as a primary responsibility: in the person’s title, their listed subject expertise, or the description of their role. Subject librarians who cover GIS incidentally among many other areas (e.g., a science librarian whose subject list includes “Geography & GIS”) are not counted. Similarly, student workers and temporary positions are noted separately.
Patterns: who has GIS librarians?
By institution size
Show code
size_summary <- inst %>%group_by(size) %>%summarise(n =n(),n_with =sum(has_gis),pct = n_with / n,.groups ="drop" )p <- size_summary %>%ggplot(aes(x = size, y = pct, fill = size,text =paste0(size, "<br>", n_with, " of ", n, " institutions<br>",percent(pct, accuracy =1)))) +geom_col(show.legend =FALSE) +geom_text(aes(y = pct /2, label =paste0(n_with, "/", n)),vjust =0.5, size =3.5, color ="white", fontface ="bold") +scale_y_continuous(labels =percent_format(), limits =c(0, 1)) +scale_fill_manual(values =c("Small"="#8BB8EE","Medium"="#2774AE","Large"="#005587","Very Large"="#003B5C" )) +labs(x ="Institution Size (Carnegie)",y ="Share with dedicated GIS librarian",title ="The bigger the university, the more likely it has dedicated GIS support" ) +theme_minimal(base_size =13)ggplotly(p, tooltip ="text")
Special-focus institutions (medical schools, law schools, and other single-discipline centers) almost never have dedicated GIS librarians, which is expected. Health sciences campuses organized primarily around clinical research and graduate training tend to house geospatial support outside the library entirely when they have it at all. At UCSF, geospatial resources are administered by the Real Estate Business Intelligence unit, not the library; the primary use case is campus operations, not research consultation.
Where peer institutions are cited as health-sciences counterexamples, the detail worth checking is organizational location, not just whether GIS exists somewhere on campus. Johns Hopkins is often cited. JHU has three GIS specialists in Digital Scholarship and Data Services in the Sheridan Libraries, not in the Welch Medical Library, which routes users to those central services. What makes JHU good practice is centralized GIS expertise in the main research library that extends to serve the full campus. That’s the pattern among comprehensive R1s broadly.
The more meaningful comparison is among comprehensive doctoral universities.
Show code
type_summary <- inst %>%mutate(type_simple =case_when(str_detect(`Institutional Classification`, "Special Focus") ~"Special Focus\n(medical/law/tech)",str_detect(`Institutional Classification`, "Very Large|Large") ~"Large Comprehensive",str_detect(`Institutional Classification`, "Medium") ~"Medium Comprehensive",TRUE~"Other" ) ) %>%group_by(type_simple) %>%summarise(n =n(), n_with =sum(has_gis), pct = n_with / n, .groups ="drop") %>%arrange(desc(pct))p3 <- type_summary %>%ggplot(aes(x =reorder(type_simple, pct), y = pct,text =paste0(type_simple, "<br>", n_with, " of ", n, "<br>",percent(pct, accuracy =1)))) +geom_col(fill ="#2774AE") +geom_text(aes(y = pct /2, label =paste0(n_with, "/", n)),hjust =0.5, size =3.5, color ="white", fontface ="bold") +scale_y_continuous(labels =percent_format(), limits =c(0, 1.1)) +coord_flip() +labs(x =NULL, y ="Share with dedicated GIS librarian",title ="Large comprehensive R1s without a GIS librarian are the exception" ) +theme_minimal(base_size =13)ggplotly(p3, tooltip ="text")
Team size
Among the 82 institutions with dedicated GIS librarians, many have built multi-person teams.
Show code
team_summary <- inst %>%filter(has_gis) %>%count(team_size_cat, name ="n_insts") %>%mutate(pct = n_insts /sum(n_insts))p4 <- team_summary %>%ggplot(aes(x = team_size_cat, y = n_insts,text =paste0(team_size_cat, " person(s)<br>", n_insts, " institutions (",percent(pct, accuracy =1), ")"))) +geom_col(fill ="#2774AE") +geom_text(aes(y = n_insts /2, label = n_insts),vjust =0.5, size =4, color ="white", fontface ="bold") +labs(x ="GIS team size",y ="Number of institutions",title ="One GIS librarian is the floor at comparable institutions; many have built teams" ) +theme_minimal(base_size =13)ggplotly(p4, tooltip ="text")
Notable multi-person GIS teams:
Show code
people %>%count(Institution, name ="Staff") %>%filter(Staff >=2) %>%arrange(desc(Staff), Institution) %>%left_join(inst %>%select(Institution, size, sector), by ="Institution") %>%rename(`Institution`= Institution, `GIS Staff`= Staff,`Size`= size, `Sector`= sector) %>%kbl() %>%kable_styling(bootstrap_options =c("striped", "hover"), full_width =FALSE)
Institution
GIS Staff
Size
Sector
Washington University in St Louis
4
Medium
Private
Arizona State University Campus Immersion
3
Very Large
Public
Johns Hopkins University
3
Large
Private
Stanford University
3
Large
Private
Brigham Young University
2
Very Large
Private
Harvard University
2
Large
Private
Massachusetts Institute of Technology
2
Medium
Private
North Carolina State University at Raleigh
2
Very Large
Public
Northeastern University
2
Large
Private
Ohio State University-Main Campus
2
Very Large
Public
Rice University
2
Medium
Private
The University of Texas at Arlington
2
Very Large
Public
University of California-Santa Barbara
2
Large
Public
University of Hawaii at Manoa
2
Large
Public
University of Michigan-Ann Arbor
2
Very Large
Public
University of Minnesota-Twin Cities
2
Very Large
Public
University of Pennsylvania
2
Large
Private
University of South Carolina-Columbia
2
Large
Public
University of Utah
2
Large
Public
University of Virginia-Main Campus
2
Large
Public
University of Washington-Seattle Campus
2
Very Large
Public
Vanderbilt University
2
Medium
Private
Virginia Polytechnic Institute and State University
2
Large
Public
Research intensity and GIS support
Research expenditure, from the NSF Higher Education R&D Survey (HERD, FY2024), is a strong predictor of whether an institution has dedicated GIS support. Institutions that invest heavily in research tend to have complex, multidisciplinary data needs that GIS librarians are well positioned to support.
GIS coverage by R&D expenditure tier
Show code
tier_data <- inst %>%filter(!is.na(herd_rank_2024), herd_rank_2024 >0) %>%mutate(rd_tier =case_when( herd_rank_2024 <=25~"Top 25 (>~$1B R&D)", herd_rank_2024 <=50~"Ranks 26–50", herd_rank_2024 <=100~"Ranks 51–100", herd_rank_2024 <=200~"Ranks 101–200",TRUE~"Ranks 200+" ) %>%factor(levels =c("Top 25 (>~$1B R&D)", "Ranks 26–50","Ranks 51–100", "Ranks 101–200", "Ranks 200+")) ) %>%group_by(rd_tier) %>%summarise(n =n(), n_with =sum(has_gis), pct = n_with / n, .groups ="drop")p5 <- tier_data %>%ggplot(aes(x = rd_tier, y = pct,text =paste0(rd_tier, "<br>", n_with, " of ", n, "<br>",percent(pct, accuracy =1)))) +geom_col(fill ="#2774AE") +geom_text(aes(y = pct /2, label =paste0(n_with, "/", n)),vjust =0.5, size =3.5, color ="white", fontface ="bold") +scale_y_continuous(labels =percent_format(), limits =c(0, 1.1)) +labs(x ="HERD R&D Expenditure Tier (FY2024)",y ="Share with dedicated GIS librarian",title ="The higher the research investment, the more certain the GIS support" ) +theme_minimal(base_size =13)ggplotly(p5, tooltip ="text")
The top 25 R&D institutions
Among the 25 highest R&D-spending universities in the country, the norm is clear. Institutions at this level of research investment consistently maintain dedicated GIS support.
medical_campuses <-c("Baylor College of Medicine","Medical University of South Carolina","The University of Tennessee Health Science Center","The University of Texas Health Science Center at Houston","The University of Texas Health Science Center at San Antonio","University of California-San Francisco","University of Maryland, Baltimore","University of Nebraska Medical Center","University of Texas Southwestern Medical Center","Weill Medical College of Cornell University")scatter_data <- inst %>%filter(!is.na(herd_rd_billions)) %>%mutate(marker_group =case_when(str_detect(Institution, "Los Angeles") ~"UCLA", Institution %in% medical_campuses ~"Health sciences campus",TRUE~"Comprehensive R1" ),hover_text =paste0( Institution, "<br>HERD Rank: #", herd_rank_2024,"<br>R&D: $", round(herd_rd_billions, 2), "B<br>",if_else(has_gis, "✓ Has GIS librarian", "✗ No GIS librarian") ) )p6 <-plot_ly() %>%add_trace(data =filter(scatter_data, marker_group =="Comprehensive R1"),x =~herd_rd_billions,y =~jitter(as.integer(has_gis), amount =0.05),type ="scatter", mode ="markers", name ="Comprehensive R1",marker =list(size =8, opacity =0.65, color ="#333333"),text =~hover_text, hoverinfo ="text" ) %>%add_trace(data =filter(scatter_data, marker_group =="Health sciences campus"),x =~herd_rd_billions,y =~jitter(as.integer(has_gis), amount =0.05),type ="scatter", mode ="markers", name ="Health sciences campus",marker =list(size =8, opacity =0.8,color ="rgba(0,0,0,0)",line =list(color ="#999999", width =1.5)),text =~hover_text, hoverinfo ="text" ) %>%add_trace(data =filter(scatter_data, marker_group =="UCLA"),x =~herd_rd_billions,y =~jitter(as.integer(has_gis), amount =0.05),type ="scatter", mode ="markers", name ="UCLA",marker =list(size =11, opacity =1, color ="#2774AE"),text =~hover_text, hoverinfo ="text" ) %>%layout(xaxis =list(title ="Total R&D Expenditure FY2024 (billions USD)"),yaxis =list(title ="", tickvals =c(0, 1),ticktext =c("No GIS librarian", "Has GIS librarian")),title ="Research expenditure and GIS librarian presence across all R1s",showlegend =TRUE,legend =list(orientation ="h", y =-0.18, x =0),annotations =list(list(x =1.88, y =0.1, text ="UCLA", showarrow =TRUE,arrowhead =2, ax =30, ay =-30,font =list(color ="#2774AE", size =12)) ) )p6
Geographic distribution
The map below shows all 187 R1 institutions. Green markers indicate a dedicated GIS librarian is present; red markers indicate none was found. Marker size reflects team size. Click any marker for details.
Show code
map_data <- inst %>%filter(!is.na(lat), !is.na(lon)) %>%left_join( people %>%group_by(Institution) %>%summarise(staff_names =paste(Name, collapse ="<br>"), .groups ="drop"),by ="Institution" ) %>%mutate(color =if_else(has_gis, "#2774AE", "#D60000"),radius =case_when( team_size ==0~6, team_size ==1~8, team_size ==2~10,TRUE~13 ),popup =case_when( has_gis ~paste0("<strong>", Institution, "</strong><br>", size, " · ", sector, "<br>","<em>GIS Staff (", team_size, "):</em><br>", staff_names ),TRUE~paste0("<strong>", Institution, "</strong><br>", size, " · ", sector, "<br>","<em>No dedicated GIS librarian found</em>" ) ) )# UCLA row for special markerucla_row <- map_data %>%filter(str_detect(Institution, "Los Angeles"))leaflet(map_data) %>%addProviderTiles(providers$OpenStreetMap) %>%addCircleMarkers(lng =~lon, lat =~lat,color =~color,fillColor =~color,fillOpacity =0.8,opacity =1,radius =~radius,stroke =TRUE,weight =1,popup =~popup,label =~Institution ) %>%# UCLA highlighted with a distinct markeraddCircleMarkers(lng = ucla_lon, lat = ucla_lat,color ="#005587",fillColor ="#FFD100",fillOpacity =1,radius =10,weight =2,popup ="<strong>University of California–Los Angeles</strong><br> Very Large · Public<br> <em>GIS librarian position eliminated;<br>no current dedicated GIS librarian</em>",label ="UCLA" ) %>%addLegend(position ="bottomright",colors =c("#2774AE", "#D60000", "#FFD100"),labels =c("Has dedicated GIS librarian", "No dedicated GIS librarian", "UCLA"),title ="GIS Support Status",opacity =0.8 )
UCLA in context
University of California–Los Angeles held the #1 public university ranking (US News) through 2024, the same years it had a dedicated GIS librarian. That position was cut in 2024.
What went with it was the work Dr. Zhiyuan Yao had built: GIS instruction for courses, research consultations, workshop programming, and administration of the campus ArcGIS Online account she had grown from fewer than 100 users to over 1,100. By late 2025, with nobody managing the platform, ArcGIS Online had grown to 230–240% of its licensed storage limit. An internal OARC report to UCLA leadership identified the vacant position as a root cause. These aren’t hypothetical risks; they’re already playing out.
UCLA is the only UC campus that had a GIS librarian and eliminated the role. UC Santa Cruz never had a dedicated GIS librarian. UC Merced didn’t either, until filling a posted position in 2026. Every institution UCLA competes with directly for rankings, faculty, and research funding has a dedicated GIS librarian; several have expanded to multi-person teams.
A question worth addressing before going to the data: why the library specifically, rather than OARC or IT? IT keeps servers running. OARC supports computational workflows. What libraries do that neither unit does is describe data so it can be found, understood, and reused years after the project that created it has ended. Metadata, provenance, lifecycle documentation: this is what librarians are trained to do, and what no other campus unit is structured to provide.
The DSC’s LARIAC documentation makes the point concretely: a StoryMap covering three generations of aerial imagery, formats, access methods, and lineage from 30+ LA County departments. That’s what a researcher relies on when they return to LARIAC files five years later, or when a grad student needs to know how LARIAC 4 point clouds differ from LARIAC 6 oblique imagery. Without it, the data effectively disappears. That’s why peer institutions with large OARC-equivalent operations still maintain library GIS positions, and why Berkeley has the Geospatial Innovation Facility and two library GIS staff.
A six-year track record, then erasure
The 2024 elimination didn’t happen in a vacuum, and it wasn’t the first time UCLA lost dedicated GIS library expertise. Andy Rutkowski served as UCLA Library’s Geospatial Librarian until departing in 2017 for a Geospatial Librarian position at USC. Following his departure, UCLA Library commissioned an external assessment conducted by a Stanford University geospatial librarian to evaluate how to rebuild geospatial services. That report explicitly recommended hiring a dedicated GIS librarian, and it came with a specific warning:
“Be certain that you are prepared to scale GIS services once a program has been implemented. You WILL see incredible growth in demand very quickly, particularly with a well-designed instruction program. A single Geospatial Librarian will quickly find they are outpaced by demand. Be prepared to hire a GIS assistant within two to three years.”
UCLA hired Dr. Zhiyuan Yao as Spatial Data Science Librarian following that recommendation. The prediction came true: GIS Day attendance climbed steadily, ArcGIS Online users grew from fewer than 100 in 2019 to over 1,100 by 2022, and Dr. Yao led the push for a systemwide SSO model to keep pace with demand.
She was among the founding organizers of UC GIS Week, the annual system-wide event that became a fixture of the UC geospatial community. In the final academic year before the position was cut, she was fielding roughly 2.5 account-request emails per day, on top of research consultations, instruction, workshops, and platform administration.
Rather than hire the GIS assistant the 2018 report anticipated, UCLA eliminated the position entirely.
The rationale offered for that decision is worth addressing directly, because it shapes what a restoration argument has to overcome. Two objections have circulated: that research libraries shouldn’t maintain specialized staff, and that GIS falls outside the library’s core scope. These are worth taking seriously because they are specific claims, and because they contradict each other.
The argument that libraries shouldn’t have specialized staff runs directly against what peer institutions have chosen. Berkeley, Michigan, Minnesota, UNC, Washington, and Ohio State all have GIS specialists embedded in their libraries. These aren’t outliers operating outside normal library practice; they are the norm at research libraries at UCLA’s level. Specialization in data services, subject-specific research support, and digital scholarship is what research libraries do. A functional organizational model (the direction library restructuring typically moves) is defined by exactly this kind of deep domain expertise rather than generalist coverage.
The argument that GIS is outside the library’s scope depends entirely on how that scope is defined. If the library’s function includes research consultation, data curation, instruction, and making information findable and usable over time, then GIS expertise sits squarely within it.
The OARC report named the vacant library GIS position as a root cause of the current platform crisis, not the absence of an IT administrator or a gap in research computing. The report’s own framing is that GIS expertise in the library is what was missing. Defining the library’s scope narrowly enough to exclude GIS also excludes the specific function that OARC identified as necessary.
Year
Event
Pre-2017
Andy Rutkowski serves as UCLA Library’s Geospatial Resources Librarian
2017
Rutkowski departs for Geospatial Librarian position at USC; UCLA commissions external assessment
2018
Stanford assessor recommends hiring a dedicated GIS librarian; warns demand will outpace one person
2018–19
UCLA hires Dr. Zhiyuan Yao as Spatial Data Science Librarian
2020
Dr. Yao among founding organizers of UC GIS Week, the annual system-wide geospatial event across all UC campuses
2020–22
ArcGIS Online users grow from <100 to 1,100+; Dr. Yao serves as Marketing Lead for UC GIS Week
2022
Dr. Yao drives SSO transition; UCLA was the last UC campus on the old model
2023–24
74 documented consultations in the final academic year across 20+ campus departments
August 2024
Position eliminated
2025
ArcGIS Online at 230–240% of licensed storage capacity; OARC report names vacant GIS librarian as root cause
The University of California system
UCSF is a health-sciences campus with no undergraduates, so its absence here isn’t notable. Of the nine comprehensive UC campuses, seven have dedicated GIS librarians. Riverside has one current person and a replacement search open. Santa Cruz has nothing. UCLA has nothing, and unlike the others, it used to.
UCLA and Berkeley have traded the top US News public university spot in recent years. Every institution in that peer group has a dedicated GIS librarian. UCLA is the only one that doesn’t.
Show code
peers <-tribble(~Institution, ~US_News_Public, ~GIS_Staff, ~Notable_Roles,"University of California-Los Angeles", "#1 (recent yrs)", 0, "Position eliminated","University of California-Berkeley", "#1 (recent yrs)", 2, "GIS & Map Librarian; Geospatial Data Specialist","University of Michigan-Ann Arbor", "#3", 2, "Map & Geospatial Data Librarian; Librarian for Geospatial & Numeric Data","University of North Carolina at Chapel Hill", "#4", 1, "GIS Librarian","University of Virginia-Main Campus", "#4", 2, "GIS Librarian; GIS Specialist","University of California-San Diego", "#6", 1, "GIS and Map Librarian","University of Wisconsin-Madison", "#8", 1, "Map and Geospatial Data Librarian","University of Washington-Seattle Campus", "#9", 2, "Geospatial Data & Maps Librarian; GIS Infrastructure Specialist","Ohio State University-Main Campus", "#17", 2, "Geospatial Information Librarian; GIS Specialist","University of Minnesota-Twin Cities", "#19", 2, "Map & Geospatial Information Librarian; Associate Director BTAA GIN","University of Maryland-College Park", "#20", 1, "GIS and Data Librarian (GISP®)","University of Florida", "#5", 1, "GIS Librarian","Penn State University", NA_character_, 1, "GIS Specialist",) %>%mutate(Status =case_when( GIS_Staff ==0~"✗ Position eliminated",TRUE~paste0("✓ ", GIS_Staff, " staff") ),US_News_Public =replace_na(US_News_Public, "—") )peers %>%select(Institution, US_News_Public, Status, Notable_Roles) %>%rename(`US News Public Rank`= US_News_Public,`GIS Support`= Status,`Roles`= Notable_Roles ) %>%kbl(escape =FALSE) %>%kable_styling(bootstrap_options =c("striped", "hover"), full_width =TRUE) %>%row_spec(which(peers$Institution =="University of California-Los Angeles"),bold =TRUE, background ="#fef9e7")
Institution
US News Public Rank
GIS Support
Roles
University of California-Los Angeles
#1 (recent yrs)
✗ Position eliminated
Position eliminated
University of California-Berkeley
#1 (recent yrs)
✓ 2 staff
GIS & Map Librarian; Geospatial Data Specialist
University of Michigan-Ann Arbor
#3
✓ 2 staff
Map & Geospatial Data Librarian; Librarian for Geospatial & Numeric Data
University of North Carolina at Chapel Hill
#4
✓ 1 staff
GIS Librarian
University of Virginia-Main Campus
#4
✓ 2 staff
GIS Librarian; GIS Specialist
University of California-San Diego
#6
✓ 1 staff
GIS and Map Librarian
University of Wisconsin-Madison
#8
✓ 1 staff
Map and Geospatial Data Librarian
University of Washington-Seattle Campus
#9
✓ 2 staff
Geospatial Data & Maps Librarian; GIS Infrastructure Specialist
Ohio State University-Main Campus
#17
✓ 2 staff
Geospatial Information Librarian; GIS Specialist
University of Minnesota-Twin Cities
#19
✓ 2 staff
Map & Geospatial Information Librarian; Associate Director BTAA GIN
University of Maryland-College Park
#20
✓ 1 staff
GIS and Data Librarian (GISP®)
University of Florida
#5
✓ 1 staff
GIS Librarian
Penn State University
—
✓ 1 staff
GIS Specialist
One detail in that table is worth sitting with. Several of these institutions also operate dedicated campus GIS centers alongside their library staff. Berkeley has the Geospatial Innovation Facility (GIF). NC State has the Center for Geospatial Analytics. Harvard has the Center for Geographic Analysis. These are substantial, well-resourced operations. And none of them replaced the library GIS position. Berkeley has the GIF and still employs two library GIS staff.
That’s because the two functions don’t do the same thing. A campus GIS center runs research projects, trains specialists, and builds tools. Library GIS staff support researchers across every department who aren’t GIS specialists, provide data access and curation, consult on projects that cross disciplinary lines, and maintain the documentation (metadata, provenance, format guides) that makes spatial data usable years after a project ends. Peer institutions treat these as complementary because they are.
UCLA has neither a dedicated campus GIS center nor library GIS support. It’s the only institution in this peer group running without either.
UCLA’s research profile
UCLA isn’t a place where GIS is marginal. Spatial methods run through multiple major programs at once, exactly the combination that drives multi-person GIS teams at comparable institutions:
Program / Unit
GIS Relevance
Department of Geography
Core GIS and spatial analysis research and instruction
Luskin School of Public Affairs, Urban Planning
Land use, transportation, housing, spatial equity
Institute of the Environment and Sustainability
Environmental modeling, remote sensing, land cover
Digital Humanities Program
Spatial humanities, historical mapping, story maps
Dept. of Ecology and Evolutionary Biology
Species distribution models, landscape ecology
Dept. of Earth, Planetary, and Space Sciences
Remote sensing, terrain analysis
Lewis Center for Regional Policy Studies
Regional spatial data analysis
These programs aren’t producing abstract academic work. They’re generating spatial analysis that feeds into state law, government commissions, and community accountability, and most of it runs on data infrastructure the library was actively supporting.
LARIAC (the high-resolution aerial imagery and LiDAR dataset the DSC manages) runs through multiple active research programs:
Heat equity (V. Kelly Turner, Luskin Center): The National Shade Map, now covering 360+ cities nationally, derives shade measurements from LARIAC LiDAR point clouds. A 2025 npj Natural Hazards paper used LARIAC 2020 LiDAR to produce shade rasters for South LA neighborhoods, processing point cloud data through CloudCompare and ArcGIS.
Urban forestry (Thomas Gillespie and Jonathan Ocón, Geography): LARIAC land cover and LiDAR-derived canopy data used for county-wide tree species mapping, developed through a direct UCLA-LA County collaboration on urban forest planning.
CalHeatScore: Mandated by Assembly Bill 2238, this tool layers LARIAC spatial data against a decade of emergency department records to produce daily heat-risk scores for every California ZIP code. Used by CalEPA, the Department of Public Health, and Cal OES.
The library’s support extended into research most people wouldn’t associate with a map:
Million Dollar Hoods (Kelly Lytle Hernandez, Bunche Center): Combines LAPD arrest data, county booking records, and Census neighborhood data to show where nearly $1 billion in annual jailing costs concentrates across Los Angeles. DSC co-taught a course with Hernandez and has worked with the project over time.
Carceral Ecologies: Integrates EPA ECHO violation data with prison facility coordinates to document chemical and radiological hazards at carceral sites. Spatial datasets published with DOIs on UCLA Dataverse; the library’s repository infrastructure is what makes the data citable, preserved, and findable.
Center for Neighborhood Knowledge (Paul Ong): Used Caltrans shapefiles, CalEnviroScreen, and ACS data to document how freeway routing in Stockton, Pacoima, and Pasadena split and isolated communities of color. Now in front of state transportation policy via a 2025 report and published StoryMaps.
LA fire recovery: After the January 2025 fires, UCLA spatial analysis for the Blue Ribbon Commission found that corporate buyers had acquired 49% of post-fire parcel sales in Altadena, up from 10% the prior year.
Gabriel Okin’s lab in Geography runs land cover classification across the South Coast Air Basin using NAIP imagery and Sentinel-2, with workflows spanning Google Earth Engine and ArcGIS Pro. That kind of pipeline (moving between platforms, handling large rasters, navigating data access) is exactly what researchers at every career stage need support with.
This is the range: heat policy, carceral geography, environmental justice, remote sensing, disaster land analysis. All of it cycling through the library’s GIS infrastructure at different points.
Michigan, Minnesota, UNC, Washington, and Ohio State all have this same mix of programs. All of them have built teams of two or more GIS librarians. UCLA had one and cut it.
Evidence of demand: what the position actually did
DSC consultation records show the position averaged 72-82 booked consultations per year. In the last academic year before the cut (September 2023 through May 2024) there were 74 documented consultations. That count excludes walk-ins, email questions, classroom sessions, and the ArcGIS platform work running in the background.
The category breakdown confirms the position was doing more than map-making. Geospatial analysis was the core, but nearly as many consultations involved data analysis, coding support, and data wrangling - skills that only matter once a researcher has a spatial dataset and needs to do something with it.
That mix reflects a design choice. The position was titled Spatial Data Scientist specifically because it was embedded in the Data Science Center, not housed separately in a maps collection. The DSC already supported researchers with R, Python, data cleaning, and statistical methods. When a patron’s data work reached the point where it needed a spatial dimension (adding coordinates, joining to a shapefile, building a choropleth), the GIS specialist was in the same room.
The handoff went the other direction too: someone who came in with a GIS question often needed the underlying data pipeline fixed first, and DSC staff could take that part. The two functions covered each other’s gaps in ways that don’t happen when GIS lives in a separate unit. That complementarity is what the job title was meant to signal, and what gets lost when the position disappears.
A few examples show what that looked like in practice:
Wolf reintroduction and river morphology (Geography): A researcher needed help identifying spatial techniques for tracking planform changes in Yellowstone river channels after wolf reintroduction, a methodology question at the intersection of ecology, geomorphology, and spatial analysis.
Air pollution exposure modeling (Medicine): A physician researcher had lung transplant patient data geocoded to ZIP code level and needed to build ambient air pollution exposure models. ZCTAs are polygons; the exposure data was point-based. Resolving that spatial mismatch took one consultation hour and specialized knowledge no other campus unit was set up to provide.
LAPD helicopter surveillance (Public Affairs): A researcher needed both the mapping and the statistics: plotting flight trajectories, then testing whether altitude and time in orbit correlated with neighborhood characteristics. The data was public record. Knowing what to do with it required GIS expertise.
The rest followed the same pattern: environmental scientists working with LARIAC LiDAR to map impervious surfaces in Santa Monica, urban planners joining census tract data to parcel shapefiles to track gentrification, digital humanities faculty embedding StoryMaps into linguistics courses, epidemiologists building spatial lag models for disease incidence at the zip code level.
This wasn’t a GIS specialist serving a handful of geography grad students. It was research infrastructure spanning medicine, public health, ecology, social science, and the humanities - at an institution that generates $1.5 billion in annual research expenditure.
The consultation data captures one-on-one support but not the infrastructure work running alongside it. The position managed UCLA’s access to LARIAC (the high-resolution aerial imagery and LiDAR dataset) through the Data Science Center. The DSC still provisions accounts through EagleView and maintains the data assets in storage. What it can no longer provide is the domain expertise that makes those assets usable.
A GIS librarian handles that side: explaining what format a file is in, how LiDAR point clouds get converted to elevation models, what the difference is between the DSM and DEM products, which LARIAC vintage is appropriate for a given research question. Researchers can still get an account. They just can’t get help knowing what to do with what they find.
The DSC also provided computing infrastructure that sits in a gap most campus units don’t address. A dedicated GIS workstation and a deep learning machine, both accessible remotely, gave researchers a place to run jobs that a laptop can’t handle but don’t justify a slot in an HPC queue. Classifying aerial imagery, processing large LiDAR datasets, running spatial deep learning models: this is mid-tier work that shows up constantly in geospatial research.
OARC handles the computationally intensive end. Researchers handle the light end on their own machines. The DSC covered the middle, with equipment maintained specifically for this kind of work and staff who knew how to use it. We are not aware of equivalent infrastructure in other library units at UCLA.
GIS instruction: courses, workshops, and campus reach
The consultation record documents one-on-one support. The instruction record shows what grows around it when the role has time to develop.
The most striking example: in Winter 2021, the GIS librarian co-taught EH207: Introduction to Geographic Information System with Prof. Michael Jerret in the Fielding School of Public Health - 8 lab sessions, 19 students, curriculum developed for the course. Most library GIS positions operate at the edges of instruction, guest lecturing into courses run by departments. At UCLA, the position had grown to the point of being embedded in a for-credit graduate course, developing its technical lab content from scratch.
Course-integrated guest instruction ran across multiple departments:
SPAN/LING 191A: The Sounds of Spanish, English, and Spanglish in LA
Spanish and Portuguese
10
Race-Making and Geography
Chicana/o and Central American Studies
14
Library workshops ran continuously alongside the course work. Between 2020 and 2023, Dr. Yao taught or helped at workshops reaching well over 400 attendees in UCLA-sponsored sessions alone - QGIS fundamentals and raster analysis (110+ attendees), Data Carpentry geospatial R (32), geospatial analysis with R (multiple sessions), StoryMaps (33), and R and Python workshops. She held a Carpentries instructor certification and served as a certified instructor and helper across more than a dozen workshops.
The system-level contribution went beyond workshops. Dr. Yao was among the founding organizers of UC GIS Week, the annual event that grew into the primary gathering for geospatial practice across the UC system, drawing participation from all campuses. She served as Marketing Lead for the inaugural cycles in 2020 and 2021, and remained on the UC GIS Leadership group through the end of her tenure. Beginning in 2022, she also took on a UC-system-level role as an organizer for UC-wide workshops. In the 2023 cycle alone:
Workshop
Attendees
The Unix Shell
164
Plotting and Programming in Python
139
R for Reproducible Scientific Analysis + Git
136
Tidy Data for Librarians
95
SQL
83
That’s 617 people across the UC system in a single workshop cycle, organized out of UCLA’s library.
None of this has a clear home now. The Carpentries program still runs at UCLA, but the GIS-specific instruction - the QGIS workshops, the geospatial R sessions, the course-integrated teaching in Public Health and the humanities - left with the position. We are not aware of any unit that has resumed it.
The ArcGIS Online crisis
Here’s how the platform got to that point. In 2022, Dr. Zhiyuan Yao, working with colleagues in the Library and research computing, pushed through a proposal to move ArcGIS from a fragmented, cost-recovery model to a campus-wide single sign-on (SSO) system. Before the change, 66 departments each managed their own licenses. Students had to email an OIT staff member and wait for manual account creation. Dr. Yao was handling roughly 2.5 account-request emails per day on top of her other work, and the 2022 proposal explicitly noted that the one-admin model was “unsustainable.”
The SSO transition worked. ArcGIS Online went from fewer than 100 active users in 2019 to over 900 in 2021 and 1,145 by April 2022, with growth accelerating after the SSO launch. The site license ($15,000/year) was centralized. The friction was gone. Usage kept climbing.
Then the GIS librarian position was eliminated in 2024. The platform didn’t break because of a technical failure. It broke because the role that administered it was removed. Platform governance, user lifecycle management, and retention policy were never transferred to another unit when the position was cut. They simply stopped happening.
An internal OARC report prepared for the Executive Director of Research Computing put it plainly:
“Without formal action, ArcGIS risks possible service disruption, degraded performance, and potential charges under Esri’s institutional terms.”
The same report named the root cause:
“The absence of a GIS Librarian, formal user management policies, and governance has led to an ad-hoc administrative environment, resulting in unmanaged growth and uncertainty about data preservation, access, and long-term sustainability.”
One of five root causes listed: “Vacant GIS support roles (GIS Librarian).”
The short-term fix on the table is buying more Esri storage: $2,700/month, $32,400/year. Add that to the existing $15,000 site license and UCLA is already at $47,400/year just to maintain a degraded version of the platform, with no governance, no support, and no instruction. User lifecycle management, data preservation, research consultation, classroom training: none of that comes back with a storage purchase.
A GIS librarian costs $80,000–$115,000/year at UC. For roughly twice the cost of the current broken state, the Library gets the platform actually managed, plus research support, instruction, and the expertise to prevent the next storage crisis. The OARC report noted that the Library “previously provided comprehensive GIS support through a dedicated GIS librarian position that included platform administration, research consultations/partnerships, and classroom instruction” and that without it, day-to-day platform support and technical training are simply “not available.”
Based on the UC-wide GIS librarian data collected for this report, every other comprehensive UC campus with active ArcGIS adoption has designated library or research computing staff with GIS responsibilities. UCLA does not, and the consequences are showing.
There is coordinated action across the UC system to renegotiate the Esri license terms, a recognition that the current storage and cost structure isn’t working. Nothing has been decided. OARC has stepped up to take on some platform responsibilities in the interim, which matters.
But platform access and platform support are different things. Negotiating better license terms lowers the barrier to entry. It does not restore the support layer: the consultations, the instruction, the data management, the domain expertise that makes the tools useful once someone has access to them. A researcher who can log into ArcGIS Online but doesn’t know how to process a LiDAR point cloud is no better off than before. That gap is what a GIS librarian fills, and it remains unfilled.
WarningWhat happens if nothing changes
A renegotiated Esri license would address access. OARC stepping in addresses some platform operations. Neither restores the support layer. Without someone whose job is GIS expertise - not platform administration as a side task, but research consultation, instruction, data management, and domain knowledge as a primary responsibility - the same gap persists. Better license terms mean more researchers can log in. It does not mean they get help once they do.
Beyond ArcGIS:
LARIAC data becomes functionally inaccessible as domain expertise concentrates at other institutions. Researchers can get accounts; they can’t get help.
New spatial datasets produced by UCLA researchers are no longer receiving the metadata, provenance documentation, and curation support that makes them findable and reusable over time. Datasets already published (like those on UCLA Dataverse) remain accessible; the question is what happens to work being done now, without a librarian managing the data lifecycle.
UCLA continues to fall further behind peers who are building out, not cutting back. Michigan, Minnesota, Washington, and Ohio State all have two-person GIS teams. The gap widens each year.
The next storage crisis won’t announce itself in advance. Whatever data infrastructure UCLA adds without a librarian to govern it replicates the same conditions.
The cost of the position is roughly twice the annual cost of the current broken state. The storage purchase does not address any of the root causes the OARC report identified. Restoring the position addresses all of them, and brings back the research support, instruction, and platform governance that disappeared with the 2024 cut.
Methods and data notes
Data collection period: February–March 2026
Source: Institutional library staff directories, subject librarian pages, and GIS service pages at all 187 Carnegie R1 institutions
Inclusion criteria: Primary GIS/geospatial/maps responsibility as evidenced by title or role description. Incidental GIS coverage by subject librarians not counted.
Caveats: Staffing changes rapidly. This represents a snapshot; positions may have been filled or vacated since collection. Institutions with vacant GIS positions are coded as “no dedicated GIS librarian” unless a named person was confirmed.
Carnegie classification source: 2025 Carnegie Classification of Institutions of Higher Education
Consultation data: The DSC consultation records analyzed here come from a booking system (Calendly/Acuity-based) used by Dr. Zhiyuan Yao from approximately 2020 through June 2024. Records include appointment date, requester department, and appointment reason/tags. Rows were filtered to Yao bookings and tagged service categories.
This captures booked one-on-one consultations only. Excluded:
Internal meetings, interviews, and DataSquad sessions
Walk-in questions and email-only interactions
Classroom sessions and platform administration work
The 2024 data runs through June; the position was vacated in August 2024. Consultation data from August 2024 onward is not available; no equivalent booking system is in active use for GIS consultations.
Author note
This report was produced by Tim Dennis, Director of the UCLA Library Data Science Center. The DSC is the unit that housed the GIS librarian position and where Dr. Zhiyuan Yao was based during her tenure. The author has a direct institutional interest in the recommendation and has tried to make that relationship explicit here rather than obscure it. The consultation data, peer comparisons, and OARC findings cited throughout are primary sources that readers can verify independently.
