Global Tree Inventory & Carbon Data System
DendroGeo turns field measurements of individual trees (circumference at 1.30 m, derived DBH = circumference/π, height, species, GNSS position, photo evidence) into allometric carbon-stock estimates (Chave et al. 2014) and 10 m land-cover context for urban parks — with moderator approval, geofencing, Monte-Carlo uncertainty and fully disclosed QA/QC in reproducible scientific reports (DGR).
Open the app · select EN for English Read the methods🌳 Individual-tree inventory
Every record carries point ID, species (canonical dictionary with sourced wood density ρ), raw circumference at 1.30 m, derived DBH, height, photo evidence and GNSS position. Raw girth_cm is preserved; dbh_cm = girth_cm / π. Allometry uses the unrounded derived DBH, while user-facing DBH is displayed to one decimal place.
⚖️ Carbon with disclosed uncertainty
AGB = 0.0673·(ρ·D²·H)^0.976; roots = 0.26·AGB; carbon = 0.47·biomass. Monte-Carlo (correlated model error, CV 22%) yields 95% confidence intervals per report.
🛰️ 10 m land-cover context
ESA WorldCover tiles are intersected with the park polygon (area-weighted border cells), with an area-balance QA gate (≤0.5%) before publication.
📄 DGR scientific reports
Each publication is an immutable, hash-verified document: metadata (DataCite 4.7-aligned), analysis fingerprint (engine, git commit, EPSG), QR to the permanent URL, citation & retraction notices. Author = the data contributor.
How is tree carbon stock calculated?
DendroGeo derives DBH from trunk circumference measured at 1.30 m above ground, then estimates biomass and carbon stock using tree height and wood density. The result is a stock estimate, not annual carbon sequestration or a personal carbon footprint. Model assumptions and measurement uncertainties are explained in the scientific methods.
Build a tree inventory with GPS position, species, circumference and height. Examine these records within the park boundary and five surface classes using urban park analysis, then export CSV, GeoJSON or QGIS files for your research workflow.
Urban park land-cover analysis
The accepted park surface uses five classes: green space, hard surface, buildings, water and bare ground. Pools and ornamental pools belong to water. ESA WorldCover raster results, Sentinel-2 spectral observations, OpenStreetMap geometries and user-accepted corrections are tracked separately. Relative NDVI describes the distribution within the same park; it is not a canopy-cover measurement. Read the scientific methods and limitations before interpreting results.
Software citation and research outputs
Şirin, N. & Şirin, S. (2026). DendroGeo: Küresel Ağaç Envanteri ve Karbon Veri Sistemi (Version 3.0.1) [Software]. Zenodo. doi:10.5281/zenodo.22948773.
This DOI identifies the software release. Cite each dataset or DGR report using its own publication metadata and DOI, when assigned. Browse the DGR publication directory for available reports and publication status.
The application has an EN language option. Translation of some information pages is still in progress. Founders: Nagihan Şirin & Sinan Şirin · Software code: MIT · approved DendroGeo data and original editorial content: CC BY-NC 4.0. Third-party materials remain under their own licenses; the DendroGeo name and logo are not licensed by these notices. See the repository for details.