← Search Investigations Case Portal: Home· Insights· Social Media· Map· Timeline· Briefing· Family Update· Subpoenas
Search Investigations
Search Investigations
Chelsea Grimm Case Portal
Confidential — Law Enforcement Sensitive

Chelsea Grimm — Missing Person Search Report

Case #CG-2023-0930  |  Coconino County Sheriff's Office  |  Report Date: April 2, 2026  |  CalTopo: caltopo.com/m/365UJR6
Cumulative search effort since December 2025
14
Search Days
259
Tracked Routes
251
Map Markers
84,000
Drone Still Images
100+
Drone Videos

Case Summary

Chelsea Grimm, age 32, was last confirmed seen on September 30, 2023 in the Kaibab National Forest, Coconino County, Arizona. Her vehicle (white Ford SUV, CA plate 8TDG491) was found on FSR 6 — out of gas, two blown tires, locked. She departed on foot carrying a tarp, backpack, ~1 gallon water, AAA almanac, bank cards, and a book. On October 20, 2023, a blue shelter matching one visible in her Instagram story was drone-spotted on a ridgeline above MK Tank, 2.19 mi south of her vehicle — believed but not confirmed to be Chelsea. Zero financial or social media activity since September 30, 2023.

This report covers Search Investigations' search operations, methodologies, technology, coverage analysis, and recommended next steps. For the full investigative briefing, case narrative, digital investigation, and phone/SIM activation analysis, see the Investigative Briefing and Key Insights pages.

Search Operations

14 Days Cumulative • Dec 2025 – Apr 2026

Search Investigations' cumulative effort since December 2025 now totals 14 field days across multiple operations plus ongoing drone image analysis. The day-by-day breakdown below is the 3/27/26 – 4/2/26 operation (5 field days + 1 analysis day); cumulative totals across all operations are summarized in the table that follows. All search areas were defined using the Lost Person Behavior model (Robert Koester, International Search & Rescue Incident Database / U.S. Department of Justice), CalTopo terrain analysis, and evidence-based behavioral profiling of the subject.

March 27 Day 1 — Initial Ground Reconnaissance & Ridge Search

Deployed ground teams to the primary search area centered on the blue shelter sighting location above MK Tank (35.37961, -112.38696). Conducted initial reconnaissance of the ridge where a blue shelter/tarp believed to be Chelsea's was spotted by drone (matching the blue shelter visible in Chelsea's backpack in her Instagram story). Covered surrounding terrain on foot, documented conditions, and began establishing search grid and GPS tracking for systematic coverage.

Tracks logged: 10 ground tracks recorded to CalTopo.

March 28 Day 2 — Expanded Ridge Coverage & Drone Deployment

Expanded search coverage across the full ridge system and surrounding terrain. Deployed DJI M30T drone for aerial survey operations, capturing high-resolution still imagery and video at multiple altitude profiles. Began systematic documentation of terrain features, water sources, and potential travel corridors from the shelter location.

Tracks logged: 11 ground tracks. Initial drone imagery collected.

March 29 Day 3 — Drone Image Analysis & Search Plan Development

Dedicated analysis day focused on reviewing all collected drone imagery, formulating the next phase search strategy based on terrain analysis, coverage gaps, and behavioral modeling. Cross-referenced drone footage with CalTopo mapping data to identify priority areas for the following search days.

March 31 Day 4 — Targeted Search Operations

Executed targeted search pattern based on Day 3 analysis. Ground teams covered washes and descending ridgelines south of the shelter location. Continued drone survey flights at varying altitudes to capture additional coverage areas. Covered the entire ridge and surrounding areas. While no physical remnants were located, the team identified the exact rock where Chelsea had been sitting based on the drone imagery alignment. From this vantage point, her full line of sight was documented — an approximately 180-degree view of the surrounding terrain.

Tracks logged: 12 ground tracks + 4 drone flights recorded.

Key observation: The position she chose offered a wide, commanding view of the terrain — a very intentional and tactical choice. From that vantage point, she would have been able to clearly see search and rescue crews operating in the area. This is consistent with her Anasazi training (selecting elevated positions with visibility) and her pattern of avoiding detection and refusing help.

April 1 Day 5 — Structured Drone Survey & Expanded Coverage

Flew structured drone patterns at approximately 400 ft AGL, moving at higher speeds while capturing rapid high-resolution imagery over a significantly expanded search radius. Knowing the exact color of Chelsea's tarp (teal/blue-green), this method allowed efficient scanning of a much broader area within a 2–3 hour operational window. Ground teams began moving downward from the ridge, covering the wash system and descending ridgelines. This area branches into three separate paths, creating a wide zone with multiple potential routes.

Tracks logged: 14 ground tracks + 2 drone survey flights. Thousands of high-resolution still images captured.

April 2 Day 6 — Continued Search & Coverage Expansion

Continued ground search operations expanding coverage southward along identified travel corridors. Additional drone flights targeting remaining coverage gaps. All imagery collected for ADIAT color detection processing.

Tracks logged: 15 ground tracks + 1 drone flight recorded.


Cumulative Search Effort

Metric Count Notes
Total Ground Tracks 259 Logged via CalTopo across all operations (Dec 2025 – Apr 2026)
Search Days 14 Cumulative field days across all operations (Dec 2025 – Apr 2026)
Drone Flights 18+ DJI M30T, multiple altitude profiles (67m – 400ft AGL)
High-Res Still Images 84,000 Captured cumulatively across all drone flights for analysis
Video Files 100+ Terrain documentation, flyovers, targeted surveys
Map Markers 251 Waypoints, evidence, structures, water sources, anomalies
Homes/Structures Mapped 160+ All structures within search radius documented
ATV Search Tracks 4 Extended range reconnaissance on forest roads
Water Tanks Checked 6 MK, Frenchman, Paradise, Mesa, Toughy, Elk

Technology & Detection Systems

AI-Assisted Analysis

🤖 ADIAT — Color Detection Active

ADIAT (Autonomous Detection of Items in Aerial/Terrestrial imagery) is an AI-powered detection system that scans drone imagery for specific color signatures using a Matched Filter algorithm. For this case, the system was configured to detect the exact hex color of Chelsea's tarp/shelter:


Target Color: RGB (22, 111, 151)
Match Filter Threshold: 0.5 | Algorithm: MatchedFilter

Status: Initial run completed on April 1 batch. Detected 2 areas of interest with confidence scores of 50.8% and 41.7%. All 84,000 cumulative drone images (since Dec 2025) are currently being processed through ADIAT for comprehensive color detection across the full search area.


Detection example: The system flagged a small teal/cyan object partially obscured by tree canopy at coordinates 35.3637, -112.3738 — demonstrating its capability to isolate color-matched items within dense vegetation from drone altitude. Note: this same coordinate is the “Under Review” ADIAT hit documented in Key Findings below; the object has not yet been ground-verified.

👁 Eagle Eye — Anomaly Detection Active

All 84,000 high-resolution still images and 100+ videos (cumulative since Dec 2025) have been processed through Eagle Eye, an anomaly detection platform that identifies visual irregularities in aerial and ground-level imagery. The system flags potential items of interest including:

  • Unnatural geometric shapes in natural terrain
  • Color anomalies inconsistent with surrounding environment
  • Man-made objects, shelters, or disturbances
  • Linear features or trails not matching known paths

Status: Initial review pass complete. All flagged anomalies are being cross-referenced with ADIAT color detection results. Imagery will be double-reviewed with both systems to ensure no items of interest are missed.

🗺 CalTopo — Search Management Active

All search operations are mapped and tracked in real-time via CalTopo (caltopo.com/m/365UJR6). The map includes:

  • 259 GPS-tracked search routes with full breadcrumb trails
  • 251 markers: evidence, water sources, structures, anomalies
  • Probability heatmap (606-cell weighted grid)
  • Coverage gap analysis overlay
  • Distance rings at 1/2/3/5/8 mi from shelter
  • Water dependency mapping
  • 109 team-sized search assignment segments
  • 7 KML overlays for different analytical views

📡 DJI M30T Drone Platform Deployed

Primary aerial platform for search operations. Capabilities deployed:

  • Wide-angle camera: High-resolution stills at 2.41–2.45 cm/px ground sample distance
  • Variable altitude profiles: Low passes (67m/220ft) for detail, high passes (120m/400ft) for broad area coverage
  • Structured survey patterns: Systematic grid flights for complete area coverage
  • GPS-tagged imagery: Every frame geolocated with drone position, altitude, and orientation for precise AOI coordinate estimation

Search Methodology

Evidence-Based

Lost Person Behavior Model

All search areas were defined using the Lost Person Behavior framework established by Robert Koester and published through the International Search & Rescue Incident Database (ISRID), referenced by the U.S. Department of Justice. This model is the industry standard used by virtually every organized search and rescue team in the United States and provides statistically validated probability zones based on subject category, terrain, and environmental factors.


Chelsea's profile was classified as follows under standard Koester / ISRID LPB framework:

  • Primary category: Despondent (intentional, seeking isolation)
  • Overlay: Mental Illness — religious/spiritual psychosis
  • Modifier: Wilderness-trained (Anasazi Foundation, 49-day primitive survival program)
  • Terrain: Mountain / High-Desert Pinyon-Juniper

Key parameters from the model that shaped our search strategy:

  • Point Last Seen (PLS): Vehicle location at 35.40904, -112.40133 on FSR 6
  • Initial Planning Point (IPP): Possible blue shelter sighting on ridge at 35.37961, -112.38696 (2.19 mi S of PLS)
  • Statistical distance bands: 25th/50th/75th/95th percentile rings applied from both PLS and shelter
  • Terrain-weighted probability: Higher weighting given to ridgelines, drainages, and water sources consistent with trained behavior
  • Behavioral modifiers: Subject's Anasazi survival training, intentional disappearance indicators, and confirmed southward movement pattern all factored into probability distribution
1
Evidence Review
Laptop, phone, social media, receipts, witness statements, bodycam footage, prior case files
2
Behavioral Profiling
Lost Person Behavior model, Anasazi training analysis, psychosis assessment, movement pattern analysis
3
Terrain Analysis
CalTopo mapping, water source network, ridgeline/drainage mapping, travel corridor identification
4
Field Operations
Systematic ground search, structured drone surveys, GPS tracking, real-time gap analysis
5
AI Detection
Eagle Eye anomaly detection, ADIAT color matching, double-review of all imagery with detection software

Key Findings

⛺ Blue Shelter / Tarp Sighting — Primary Anchor Point Unconfirmed

A blue shelter / teal tarp was spotted by drone on a ridgeline directly above MK Tank at 35.37961, -112.38696 — matching the blue shelter visible in Chelsea's backpack in her Instagram story. This is 2.19 miles south of her vehicle on a bearing of 158°, consistent with her southward on-foot movement pattern. While this has not been confirmed as Chelsea's, the camp location is consistent with Anasazi site selection training: elevated position on a ridge 580 feet directly above a full water source.

👇 Exact Sitting Rock Identified March 31

During Day 4 operations, the team identified the exact rock where a person had been sitting, verified through alignment with earlier drone imagery of the blue shelter sighting. From this vantage point, the team documented her complete ~180-degree line of sight over the surrounding terrain.

Critical implication: From this position, Chelsea would have had clear visual contact with any search and rescue activity in the area. Her choice of this position — combined with her pattern of refusing help from every person she encountered — strongly suggests she was actively avoiding detection. This is consistent with both her Anasazi ridge-position training and her deliberate disconnection behavior.

🤖 ADIAT Detection — Color Match at 35.3637, -112.3738 Under Review

The ADIAT system flagged two consecutive drone images from the April 1 flight showing a color-matched object in vegetation approximately 1.1 miles SSE of the shelter location:

ImageGPS (Est. Ground)ConfidenceColor MatchAGL
DJI_20260401133146_0343 35.363653, -112.373804 50.8% Hue 170° #00ffd7 67.4m
DJI_20260401133149_0344 35.363554, -112.374109 41.7% Hue 185° #00e7ff 66.4m

Both images show a small teal/cyan object partially obscured by tree canopy. The detection location falls on the "Possible Path Taken" route (SSE from the shelter) and is within the Priority 1 search corridor. This location requires ground verification.

💧 Water Source Analysis Verified

The shelter sighting is directly above MK Tank, which was confirmed full at the time. With approximately 1 gallon of water capacity, Chelsea had roughly one day of hydration in October Arizona conditions (highs 60s, lows 30s °F). She would need to locate a new water source within approximately 1 day of departure or face dehydration. At a sustainable wilderness walking pace, this constrains her effective travel radius from MK Tank to roughly 6–10 miles before a required resupply.

Water SourceDistance from ShelterSearch Tracks (0.5mi)Status
MK Tank0.11 mi (580 ft)PRIMARY — confirmed full
Frenchman Tank1.02 mi E26Well covered
Paradise Tanks2.13 mi NW13Moderate coverage
Mesa Tank2.46 mi E2BARELY SEARCHED
Toughy Tank2.52 mi NE5Light coverage
Elk Tank3.46 mi NE0ZERO COVERAGE

🧭 Anasazi Foundation Connection Corroborated

Additional research uncovered accounts from a former Anasazi Foundation counselor describing how participants were trained to sustain themselves with nutrient-dense supplies such as Gatorade powder, salt, and baking soda to prevent dehydration. This directly aligns with Chelsea's Dollar General receipt, which included salt and baking soda, suggesting she was actively applying wilderness survival skills learned during her time at the program.


Additional behavioral parallels identified:

  • Tarp-based sheltering: Standard Anasazi technique; Chelsea departed with a tarp (not a tent)
  • Ridge-to-ridge movement: Frequently moving between elevated positions for visibility and tactical advantage
  • Water source proximity camping: Her shelter placement 580 ft above MK Tank is textbook Anasazi site selection
  • LifeStraw-type water filtration: Commonly used in wilderness programs (unconfirmed if Chelsea had one)

Reference: Former counselor account (Reddit)

Search Coverage Analysis

Coverage by Distance from Shelter (IPP)

0 – 0.5 mi
95%
0.5 – 1.0 mi
86%
1.0 – 1.5 mi
80%
1.5 – 2.0 mi
81%
2.0 – 2.5 mi
73%
2.5 – 3.0 mi
34%

Major gap at 2.5–3.0 miles: Only 34% coverage at this critical distance band. This zone falls along Chelsea's probable southward travel corridor and includes the transition from searched to unsearched terrain. Coverage beyond 3 miles is minimal to nonexistent.

Key Coverage Gaps

DirectionGap SizeDistanceCenter PointPriority
South 12+ cells 0.8 – 2.0 mi 35.358, -112.387 P1
South 14 cells 2.5 – 3.0 mi 35.344, -112.387 P1
East 23 cells 2.0 – 3.0 mi 35.379, -112.342 P2
West 20 cells 2.0 – 3.0 mi 35.379, -112.431 P3
Southwest 6 cells 0.9 – 1.8 mi 35.362, -112.398 P2

Areas Searched & Cleared

The following areas have been thoroughly covered through combined ground and aerial search operations:

  • All ridges, washes, and water tanks within the immediate radius of PLS/vehicle location
  • Complete 360-degree coverage within 0.5 miles of the blue shelter/tarp sighting
  • Full ridge system where the shelter was located, including identification of the exact sitting position
  • MK Tank and Frenchman Tank areas (well covered with 26+ tracks)
  • Paradise Tanks area including white truck corridor (moderate coverage)
  • Primary forest roads (FR 6, FR 124) and connecting routes

Behavioral Insights

Direction Assessment

Most probable: South to SSW (bearing 160–200°)


Supporting evidence for south:

  • Confirmed on-foot bearing averages 190°
  • I-40 is 8.6 miles south (road noise audible)
  • Bank cards suggest she intended to reach civilization
  • Egress drainage provides natural walking path SSW
  • Largest search gaps are directly south

Secondary possibility — ESE toward "Humphreys" (Humphreys Peak, ~45 mi): She told the Worthingtons this was her destination, and the mountain holds spiritual significance. However, 45 miles with a tarp and 1 gallon of water would be nearly impossible.

Recommended Next Steps

✓ COMPLETED ADIAT Processing & Ground-Verification

All 84,000 drone images (cumulative since Dec 2025) have been processed through ADIAT color detection matching Chelsea's tarp. Two areas of interest were flagged at 35.3637, -112.3738 (1.1 mi SSE of the shelter) and ground-verified. Processing and verification complete.

Priority 1   Search S/SSE Corridor, 2–5 Miles from Shelter

The 34% coverage zone at 2.5–3.0 miles on Chelsea's primary bearing is the most critical unsearched area. This corridor (35.340–35.370 lat, -112.410 to -112.360 lon) includes the CG-31 and CG-30 search segments at 3.4–3.7 miles south. With 1 gallon of water, she could reach approximately 5 miles before a dehydration crisis. Search drainages, ridgelines, and terrain traps.

Priority 1   Extend Egress Drainage Search Beyond 4-Mile Endpoint

The mapped egress route running SSW from the shelter ends at 4 miles. The drainage continues south toward I-40 (8.6 mi total). Walk the drainage channel from mile 4 to mile 7, checking both banks and side channels within 0.25 miles. Our best estimate places Chelsea approximately 6 miles SSW of the shelter (35.2960, -112.4180), along this drainage.

Priority 2   Double-Review All Imagery with Detection Software

All drone footage (84,000 stills, 100+ videos cumulative since Dec 2025) to be reviewed through both Eagle Eye anomaly detection and ADIAT color detection. This dual-system approach ensures maximum probability of identifying items of interest that may have been missed in single-pass manual review.

Priority 2   Search Mesa Tank & Elk Tank Corridor (East)

If Chelsea angled east toward "Humphreys," Mesa Tank (2.46 mi E, only 2 search tracks) and Elk Tank (3.46 mi NE, zero coverage) are the first water sources on that route. Search 1-mile radius around each tank plus the connecting corridor between them.

Priority 3   Follow Up on Outstanding Requests

  • Sheriff's Department helicopter footage: Request still pending for footage from the ridge area where a drone-spotted shelter (and possibly a figure) was recorded. Sighting not yet confirmed to be Chelsea. Follow up to obtain this corroborating evidence.
  • White truck at Paradise Tanks: Vehicle driving erratically at 5:08 PM on Sept 30, 1.55 miles from Chelsea's car, 45 minutes before her final photo. Investigate further to rule out involvement.
  • Zach McGaughey statement: Saw Chelsea's car with all doors open and someone moving a bag to a signpost. Denies seeing 4 people (contradicts original report). Resolve discrepancy.