Dendrochronology Dating Tool

A Python CLI tool for dating historic timber by cross-dating ring width measurements against ITRDB reference chronologies.

Overview

This tool helps determine when trees were felled by:

1. Extracting ring width measurements from scanned wood cross-sections
2. Cross-dating against publicly available ITRDB chronologies
3. Identifying the calendar year of the outermost ring (felling year if bark edge present)

Designed for dating historic New England timber (1600s-1800s), but works for any region with ITRDB coverage.

Installation

# Clone and enter directory
cd dendrochronology

# Create virtual environment and install
python3 -m venv venv
source venv/bin/activate
pip install -e ".[dev]"

Requires Python 3.9+.

Quick Start

# Activate environment
source venv/bin/activate

# Download reference chronologies (already done for northeastern US)
dendro download --states=me,nh,vt,ma,ct,ri,ny

# Check available references
dendro info

# Measure rings from a scan (interactive)
dendro measure sample.tiff --dpi=1200 --output=sample.csv

# Cross-date against references
dendro date sample.csv --era-start=1750 --era-end=1850 --bark-edge

CLI Commands

dendro download

Download reference chronologies from ITRDB (NOAA NCEI).

dendro download --states=me,nh,vt,ma --species=PIST,TSCA,QUAL,QURU

Options:

• --states, -s: State codes, comma-separated (default: me,nh,vt,ma,ct,ri,ny)
• --species, -p: Species codes, comma-separated (default: PIST,TSCA,QUAL,QURU)
• --output, -o: Output directory (default: data/reference)
• --overwrite: Re-download existing files

dendro info

Show information about downloaded reference chronologies.

dendro info

dendro measure

Extract ring widths from a scanned wood sample using an interactive viewer.

dendro measure sample.tiff --dpi=1200 --output=sample.csv

Options:

• --dpi, -d: Scanner resolution in DPI (default: 1200)
• --output, -o: Output file for measurements (CSV)

Interactive controls:

1. Path mode: Click to mark measurement path from bark (outer) to pith (center)
2. Press ENTER to switch to ring marking mode
3. Ring mode: Click to mark ring boundaries, or press 'A' for auto-detect
4. Press ENTER to save and close

dendro date

Cross-date a sample against reference chronologies.

dendro date sample.csv --era-start=1750 --era-end=1850 --bark-edge

Options:

• --reference, -r: Reference directory (default: data/reference)
• --era-start: Earliest possible felling year (default: 1600)
• --era-end: Latest possible felling year (default: 1900)
• --species, -p: Filter by species codes
• --states, -s: Filter by state codes
• --bark-edge/--no-bark-edge: Sample includes bark edge (default: yes)
• --output, -o: Save results to JSON file
• --top, -n: Number of top matches to display (default: 10)

dendro parse

Parse and display contents of a Tucson format file.

dendro parse data/reference/nh/nh001.rwl

Sample Preparation

Extraction Methods

Option A: Cut Section (Recommended for exposed beam ends)

• Cut a thin cross-section (1-2 cm thick) from beam end
• Provides full cross-section for scanning
• Best quality for ring measurement

Option B: Increment Borer

• 5mm diameter core from pith to bark
• Minimal damage, leaves small pluggable hole
• Core must be mounted and sanded

Option C: Wedge Sample

• V-cut into beam edge where not visible
• Good compromise between damage and quality

Critical Requirements

1. Include bark edge: Sample MUST extend to the original outer surface (bark or waney edge) for exact felling year
2. Include pith if possible: The center helps verify ring count
3. Clear radial path: Need unobstructed path from near-center to bark
4. Multiple samples: Take 2-3 from different beams for cross-verification

Surface Preparation

1. Sand progressively: 120 → 220 → 400 grit
2. For pine/hemlock: Wipe with mineral spirits to enhance contrast
3. Goal: Clearly visible ring boundaries (dark latewood vs light earlywood)

Scanning

• Minimum: 1200 DPI (workable with good preparation)
• Recommended: 2400 DPI for narrow rings
• Use TIFF format (no compression artifacts)
• Scan in color, convert to grayscale in software
• Place flat sanded surface against scanner glass
• Use black backing behind sample

How Cross-Dating Works

Algorithm

1. Detrend: Remove age-related growth trend from ring width series using cubic spline or negative exponential curve
2. Standardize: Convert to dimensionless indices with mean ~1.0
3. Slide: Move sample series along reference chronology one year at a time
4. Correlate: Calculate Pearson correlation coefficient at each position
5. Rank: Find positions with highest correlation and t-value

Confidence Metrics

Metric	High Confidence	Medium	Low
Correlation (r)	> 0.60	0.45-0.60	< 0.45
T-value	> 6.0	4.0-6.0	< 4.0
Overlap (years)	> 50	30-50	< 30

Sample Length Requirements

• Minimum: 30 rings (marginal confidence)
• Recommended: 50+ rings (good confidence)
• Optimal: 100+ rings (high confidence)

Project Structure

dendrochronology/
├── pyproject.toml              # Dependencies and config
├── src/dendro/
│   ├── reference/              # ITRDB data handling
│   │   ├── tucson_parser.py    # Parse .rwl/.crn Tucson format
│   │   ├── downloader.py       # Download from NOAA NCEI
│   │   └── chronology_index.py # Index by species/location
│   ├── imaging/                # Ring width extraction
│   │   ├── ring_detector.py    # Detect ring boundaries
│   │   ├── path_sampler.py     # Sample along measurement path
│   │   └── viewer.py           # Interactive measurement UI
│   ├── crossdating/            # Dating algorithms
│   │   ├── detrend.py          # Growth trend removal
│   │   ├── correlator.py       # Sliding correlation
│   │   └── matcher.py          # Multi-chronology matching
│   └── cli/main.py             # CLI entry point
├── data/
│   └── reference/              # Downloaded ITRDB chronologies
└── tests/                      # Test suite (43 tests)

Species Codes

Common northeastern US species:

• PIST: Pinus strobus (Eastern White Pine)
• TSCA: Tsuga canadensis (Eastern Hemlock)
• QUAL: Quercus alba (White Oak)
• QURU: Quercus rubra (Red Oak)
• PCRU: Picea rubens (Red Spruce)
• THOC: Thuja occidentalis (Northern White Cedar)

Regional Marker Years

Notable climate events visible in ring patterns:

• 1816: "Year Without Summer" - very narrow ring
• 1780s: Series of severe winters
• 1790s: Generally favorable growing conditions

New Hampshire Reference Data

For dating historic NH timber, the most relevant downloaded references are:

File	Species	Site	Cores	Years
nh001	Red Spruce (PCRU)	Nancy Brook	30	1561-1972
nh002	Eastern Hemlock (TSCA)	Gibb's Brook	25	1509-1981
nh003	Red Spruce (PCRU)	Nancy Brook	31	1610-1979
nh004	Red Spruce (PCRU)	Mt. Washington	35	1678-1976
nh005	Red Pine (PIRE)	Rattlesnake Mtn	53	1690-2008

128 tree cores from NH cover the 1780-1800 period (relevant for late 1700s construction).

Connecticut River Valley References (Walpole, NH Area)

For dating timber in the southern NH / VT border region (Walpole, Charlestown, Keene), these references are particularly valuable:

Eastern Hemlock (TSCA) - Historic Buildings

File	Site	Cores	Years	Notes
vt010	West Brattleboro Apartments	10	1570-1869	~10 mi from Walpole
vt011	West Dummerston Covered Bridge	4	1557-1847	Historic structure
vt013	Green River House	4	1568-1834	Historic building
vt009	Guilford Center Meeting House	1	1453-1834	Historic building

Natural Chronologies

File	Species	Site	Cores	Years
nh002	Eastern Hemlock	Gibb's Brook	25	1509-1981
vt005	Red Spruce	Bolton Mountain	7	1668-2005
ma017	Eastern Hemlock	Cold River	8	1650-2003
ma018	Eastern Hemlock	Alander Mountain	6	1628-1987

43+ hemlock cores from the Connecticut River Valley region are ideal for dating late 1700s construction timber.

Data Sources

Reference chronologies from ITRDB (International Tree-Ring Data Bank):

• https://www.ncei.noaa.gov/pub/data/paleo/treering/measurements/northamerica/usa/
• https://www.ncei.noaa.gov/pub/data/paleo/treering/chronologies/northamerica/usa/

Running Tests

source venv/bin/activate
pytest tests/ -v

47 tests total, including 4 real-data validation tests using ITRDB chronologies.

Algorithm Validation

The cross-dating algorithm has been validated against real ITRDB data:

Test	Species	Site	Result
nh001	Red Spruce (PCRU)	Nancy Brook, NH	Correct date (0 error)
nh002	Eastern Hemlock (TSCA)	Gibb's Brook, NH	Correct date (0 error)
nh003	Red Spruce (PCRU)	Nancy Brook, NH	Correct date (0 error)
ma015	Mixed	Historic buildings, MA	Correct date (≤5 yr error)

Run validation manually:

python scripts/validate_crossdating.py --num-tests=5

Important: Cross-dating works best when reference chronologies are from:

1. The same species as your sample
2. A geographically similar area (within ~100 miles)

Example Output

CROSS-DATING RESULTS
============================================================
Sample: basement_beam_1
Length: 87 rings
Bark edge: Yes

PROPOSED FELLING YEAR: 1789
Confidence: HIGH

Top 10 matches:
------------------------------------------------------------
1. ma012 (MA)
   Felling year: 1789
   Correlation: 0.67, T-value: 7.2
   Overlap: 87 years, Confidence: HIGH

2. nh003 (NH)
   Felling year: 1789
   Correlation: 0.61, T-value: 6.1
   Overlap: 82 years, Confidence: HIGH
...

License

MIT