Reports of I.C.E. Activity — Los Angeles & the Inland Empire

Readme — ICE Activity Map

Los Angeles & the Inland Empire  |  January 2025 – April 2026

Introduction

This map visualizes crowd-sourced reports of observed ICE (Immigration and Customs Enforcement) activity across Los Angeles, Orange, Riverside, and San Bernardino Counties. It is built on data collected by People Over Papers through our platform iceout.org, where residents, rapid response volunteers, and community organizations submit real-time reports of observed enforcement activity.

The tool is designed to make the geographic and temporal scope of immigration enforcement visible at a level of detail that no government data source currently provides. It is intended as a resource for journalists, community organizers, researchers, and elected officials seeking to understand local enforcement patterns.

This is crowd-sourced data with meaningful limitations described below. Results should be interpreted as indicative of patterns, not as a precise accounting of enforcement actions.

Navigating the Map

Sidebar Filters

• Date Range — restrict the data to a specific window of time
• Status (Rapid Response Only) — filter to reports from vetted rapid response network partners
• Activity Category — filter by the type of activity reported (e.g. Arrest/Transfer, Raid, Checkpoint)

Display Modes

• Points — each dot represents one report at its reported location
• Heatmap — a density surface showing concentrations of reports

Layers

• Census Tracts — neighborhood-level boundaries
• Electoral Districts — U.S. Congressional districts (statewide California data)
• Cities — municipal boundaries within the four counties
• Hex Tiles — uniform ¾-mile hexagonal grid

Detail Panel

Clicking any highlighted geometry opens a panel with: reports over time (ignores date filter — shows full history; still reflects other filters), hour of day breakdown, and activity category breakdown. All filters except date apply to the hour and category charts.

The Show Regional Averages button opens the same three charts for the full filtered dataset.

Data

Source

All reports come from iceout.org, operated by People Over Papers / Pueblo Project Foundation. The dataset covers approved reports from January 22, 2025 – April 17, 2026 across the four-county study area. Reports flagged as Confirmed come directly from rapid response network partners and carry a higher degree of accuracy.

Congressional District Layer

The Congressional Districts layer is the only layer using statewide California data rather than the four-county dataset. This ensures districts extending beyond the study area reflect their full statewide report counts rather than being truncated at the county boundary.

Geographic Boundaries

• Census tracts — U.S. Census Bureau TIGER/Line files, 2023, via the tigris R package
• Congressional districts — CA Congress Districts and Membership 2024–2026, California State Geoportal. gis.data.ca.gov
• City boundaries — California Incorporated Cities, California State Geoportal. gis.data.ca.gov
• County boundaries — U.S. Census Bureau via tigris

Population Data

Population figures shown in the congressional district and city detail panels come from the American Community Survey (ACS) 5-year estimates, 2022, accessed via the U.S. Census Bureau API using the tidycensus R package. Variable B01003_001 (total population) is used for both congressional districts and incorporated cities/places. Reports per capita are calculated as reports per 100,000 residents.

Hex Tile Grid

Hex tiles are generated as a uniform ¾-mile (~1,207m flat-to-flat) hexagonal grid. Only tiles containing at least one report are displayed.

hex_cell_m <- 1207
counties_union_proj <- counties_union_4326 |> sf::st_transform(3310)
bbox_exp <- sf::st_bbox(counties_union_proj)
bbox_exp["xmin"] <- bbox_exp["xmin"] - hex_cell_m
bbox_exp["ymin"] <- bbox_exp["ymin"] - hex_cell_m
bbox_exp["xmax"] <- bbox_exp["xmax"] + hex_cell_m
bbox_exp["ymax"] <- bbox_exp["ymax"] + hex_cell_m

hex_grid <- sf::st_make_grid(
  sf::st_as_sfc(bbox_exp),
  cellsize = hex_cell_m,
  square   = FALSE
) |> sf::st_sf() |> sf::st_make_valid()

hex_clipped <- hex_grid[
  sf::st_intersects(hex_grid, counties_union_proj, sparse = FALSE)[, 1], ] |>
  sf::st_transform(4326) |>
  mutate(hex_id = paste0("h", row_number()))

Methods

Data Preparation

Raw reports are loaded from a CSV export of approved iceout.org submissions. Coordinates are extracted from WKT POINT geometry. Each report is spatially joined to its census tract, congressional district, city, and hex tile using sf::st_within. Reports outside the four-county boundary are excluded via spatial join against Census Bureau county polygons.

pts_in_tracts <- st_join(pts_sf, tracts_proj |> select(GEOID), join = st_within, left = TRUE)
pts_in_cd     <- st_join(pts_sf_4326, cd_simple |> select(CD_GEOID = GEOID), join = st_within, left = TRUE)
pts_in_city   <- st_join(pts_sf_4326, cities_simple |> select(CITY_GEOID = CENSUS_GEOID), join = st_within, left = TRUE)

Front-End Architecture

The map is a fully self-contained HTML file requiring no server. All data is embedded at build time as JSON arrays and base64-encoded images. Interactivity is handled by Plotly.js (v2.32.0) with a Mapbox basemap. All filtering runs client-side with no external calls after load.

arrays_json <- jsonlite::toJSON(list(
  lat        = df$latitude,    lon        = df$longitude,
  date       = df$date_str,    confirmed  = df$confirmed_clean,
  cats       = df$activity_cats_str,
  geoid      = df$GEOID_clean, cd_geoid   = df$CD_GEOID_clean,
  city_geoid = df$CITY_GEOID_clean,
  hex_id     = df$HEX_ID_clean, hour      = df$hour
), na = "null", auto_unbox = FALSE)

Limitations

• Reporting bias — Communities with greater awareness of iceout.org are likely over-represented. This should not be understood as a uniform sample of enforcement activity.
• Temporal spikes — Increases in report volume may reflect heightened public awareness as much as an actual increase in enforcement activity.
• Location accuracy — Location data reflects where the reporter observed the incident, which may not precisely match the enforcement location. Moderators review for duplicates but cannot guarantee locational precision.
• Activity categorization — Categories are selected by the reporter and may not always accurately reflect what occurred. Multiple categories can be selected per report.
• What is not captured — Enforcement actions that were not observed, not reported, or occurred in areas with lower platform awareness are not represented. This is not a complete census of ICE operations.

Contact

For questions about the data, email us at contact@puebloproject.com.

This tool is provided for informational and research purposes. Report counts and patterns should be interpreted as indicative of enforcement trends, not as a precise record of individual enforcement actions.

Activity Category Guide

Activity categories are assigned by reporters via the report submission form on iceout.org. Reporters may select as many categories for a report as are applicable.

The following key provides explanations for the less intuitive activity category options:

Welcome to the L.A. Immigration Enforcement Activity Map

Displaying data for L.A. and the Inland Empire

Since the beginning of President Trump's second term in January 2025, activity by federal immigration agents across the United States has intensified, and Los Angeles and the Inland Empire have been some of the most consistent targets. This map is a tool for understanding what that activity looks like on the ground — where it is happening, when, and how often.

Data and Limitations

The data included here is compiled from reports to Iceout.org by the public. All reports mapped on Iceout.org and included in this dataset are confirmed by trained Pueblo Project moderators as probable ICE or CBP sightings. Moderators aren't on site when reports are made, and can't guarantee the accuracy of all reports. This map does not display all ICE activity in the studied area, but is meant to document broad patterns.

Using the Map

The map is interactive, allowing users to toggle the data being displayed. Use the filters on the left sidebar to narrow the data by date range, activity category, or to show rapid response reports only — the map will adjust accordingly.

Scrolling down the left sidebar to the Layers menu, you can view report density across census tracts, cities, congressional districts, or evenly sized ¾-mile hexagonal tiles. Within any layer, click the map to open a panel showing that location's specific report counts over time, a breakdown by hour of day, and activity categories. The city and congressional district layers also provide population information, reports per capita, and the sitting US House Representative for each district.

Access the full readme on the upper bar of the site.