# Mobile Phone GPS Mobility Data for COVID-19 Research in Hermosillo, Sonora, Mexico, 2020

**Version:** 1.0.0  
**Publisher:** Universidad de Sonora  
**Publishing organization:** Departamento de Matemáticas, Universidad de Sonora  
**CKAN organization:** Matemáticas  
**Maintainer:** Jesús Francisco Espinoza Fierro  
**Maintainer email:** jesusfrancisco.espinoza@unison.mx  
**Dataset URL:** https://datos.fi-cen.unison.mx/dataset/covid19-mobility-hmo-2020  
**License:** Creative Commons Attribution 4.0 International (CC BY 4.0)

## Overview

This dataset contains mobile phone GPS mobility records for Hermosillo, Sonora, Mexico, collected during the 2020 COVID-19 pandemic. The records were used in mathematical, epidemiological, and urban mobility studies focused on residence-occupation patterns, human mobility restrictions, and multi-patch epidemic modeling.

The dataset includes the original daily CSV files and a consolidated long-format table with two additional traceability fields: `date` and `source_file`. The consolidated analytical table is distributed in CSV and Parquet formats. Each mobility record corresponds to a device-level GPS ping generated when a mobile application with location permissions was used. The dataset is released for academic research, mathematical modeling, epidemiological analysis, urban mobility studies, reproducibility, and open scientific experimentation.

## Data coverage

- **Temporal coverage of daily source files:** 2020-09-18 to 2020-12-13.
- **Timestamp reference:** UTC.
- **Local time conversion:** Hermosillo, Sonora used UTC−07:00 throughout the study period; Sonora did not observe daylight saving time during this period.
- **Spatial coverage:** Hermosillo, Sonora, Mexico.
- **Coordinate reference system:** WGS84 geographic coordinates, EPSG:4326.
- **Number of daily source files:** 87.
- **Number of consolidated records:** 80,582,452.
- **Number of device identifiers reported in the associated study:** 306,963.
- **Primary variables:** device identifier, timestamp, latitude, longitude, gender, country, age, income level, tier1, tier2, and technical provenance fields.

The file date (`date`) is a source-file traceability field derived from the daily CSV filename. It should not be confused with the ping timestamp (`timestamp`), which records the date and time of the mobility observation in UTC.

## Data resources

| Resource | Format | Distribution | Purpose |
|---|---:|---|---|
| `data/mobility_hmo_v1_0_0.parquet` | Parquet | CKAN resource | Primary analytical resource. Recommended for computational work because of the dataset size. |
| `data/mobility_hmo_v1_0_0.csv` | CSV | Institutional direct-download URL registered in CKAN | Consolidated long-format table. This file is large and may be impractical to upload through standard web forms. |
| `data/raw_daily_csv_2020-09-18_2020-12-13.zip` | ZIP | CKAN resource | Compressed archive containing the original daily CSV files as received and organized by date. |
| `metadata/data_dictionary.csv` | CSV | CKAN resource | Variable-level codebook for all fields in the consolidated dataset. |
| `docs/methodology.md` | Markdown | CKAN resource | Methodological description of data provenance, processing, quality checks, limitations, and responsible use. |
| `docs/README.md` | Markdown | CKAN resource | Public user guide for the dataset. |
| `metadata/CITATION.cff` | CFF/YAML | CKAN resource | Machine-readable citation metadata. |
| `metadata/datapackage.json` | JSON | CKAN resource | Frictionless-style metadata package describing the resources and tabular schema. |
| `metadata/ckan_metadata.json` | JSON | CKAN resource | CKAN-oriented metadata record for institutional documentation and future API use. |
| `integrity/checksums_sha256.txt` | TXT | CKAN resource | SHA-256 checksums for file integrity verification. |
| `integrity/manifest_files.csv` | CSV | CKAN resource | Machine-readable manifest of data, documentation, metadata, and script files, including file sizes, human-readable sizes, and SHA-256 hashes. Integrity files are excluded to avoid self-referential records. |
| `scripts/build_ckan_mobility_package.py` | Python | CKAN resource | Python script used to generate the primary analytical data resources and data-resource integrity files from the original daily CSV files. |

The Parquet and ZIP resources are distributed through CKAN FileStore. The consolidated CSV is distributed through an institutional direct-download URL registered in CKAN because of file size. File integrity should be verified using the SHA-256 checksums.

## Data provenance and authorization

The original mobility data source was provided for academic research through LUMEX CONSULTORES, S.C. in the framework of the service contract signed between LUMEX CONSULTORES, S.C. and Universidad de Sonora on September 14, 2020, in Hermosillo, Sonora, Mexico.

LUMEX CONSULTORES, S.C. granted authorization for the mobility data analyzed under this framework to be used in academic research articles and to be made available as Open Data in an open access repository. Any publication, derivative dataset, software, or analytical product using these data must acknowledge LUMEX CONSULTORES, S.C. as the provider of the original mobility data source for academic purposes.

The source records are associated with mobile phone GPS pings. They were generated when users accessed specific mobile applications with location permissions enabled. The exact applications are not disclosed in this dataset.

## Responsible use

The dataset contains persistent device-level identifiers, timestamps, and GPS coordinates. These fields are necessary for scientific analysis of mobility patterns over time and space, but they must be handled as sensitive mobility information.

GPS pings should not be interpreted as deterministic evidence that a person was located at an exact physical point at an exact instant. Consecutive pings associated with the same device may exhibit short-range spatial variability over seconds or minutes, reflecting device-level positioning uncertainty, application-level recording behavior, signal conditions, and data-processing effects. Coordinates should therefore be interpreted as GPS observations rather than exact statements of physical presence.

The studies derived from this dataset did not rely on exact point-level identification of places or individuals. Instead, they aggregated mobility information to larger spatial units such as AGEBs, urban zones, or patch-based regions, where small-scale GPS variability does not materially affect the conclusions.

Users must not attempt to identify, re-identify, contact, profile, locate, or infer sensitive personal routines of any individual, household, workplace, or device owner. The dataset is released for scientific, educational, methodological, and public-interest research purposes.

## Recommended citation

Espinoza Fierro, Jesús Francisco; Montoya Laos, José Arturo; Ramírez Ramírez, Lilia Leticia; Soto Barrera, Juan Pablo. (2026). Mobile Phone GPS Mobility Data for COVID-19 Research in Hermosillo, Sonora, Mexico, 2020. Version 1.0.0. Universidad de Sonora. Data source provided by LUMEX CONSULTORES, S.C.

## Dataset creators

- Jesús Francisco Espinoza Fierro — ORCID: https://orcid.org/0000-0001-5601-3634
- José Arturo Montoya Laos — ORCID: https://orcid.org/0000-0001-6952-0109
- Lilia Leticia Ramírez Ramírez — ORCID: https://orcid.org/0000-0002-9469-0887
- Juan Pablo Soto Barrera — ORCID: https://orcid.org/0000-0001-6107-1097

## Related publications

Ramírez-Ramírez, Lilia Leticia; Montoya, José A.; Espinoza, Jesús F.; Mehta, Chahak; Akuno, Albert Orwa; Bui-Thanh, Tan. (2025). *Use of mobile phone sensing data to estimate residence and occupation times in urban patches: human mobility restrictions and the 2020 COVID-19 outbreak in Hermosillo, Mexico*. Computational Urban Science, 5, 10. https://doi.org/10.1007/s43762-025-00168-y

Akuno, Albert Orwa; Ramírez-Ramírez, Lilia Leticia; Espinoza, Jesús F. (2023). *Inference on a Multi-Patch Epidemic Model with Partial Mobility, Residency, and Demography: Case of the 2020 COVID-19 Outbreak in Hermosillo, Mexico*. Entropy, 25, 968. https://doi.org/10.3390/e25070968

## Quick start

Python example for reading the Parquet resource:

```python
import pandas as pd

path = "mobility_hmo_v1_0_0.parquet"
df = pd.read_parquet(path)
print(df.shape)
print(df.head())
```

Python example for converting timestamps to Hermosillo local time:

```python
import pandas as pd

utc_time = pd.to_datetime(df["timestamp"], utc=True, errors="coerce")
df["timestamp_hmo"] = utc_time.dt.tz_convert("Etc/GMT+7")
```

`Etc/GMT+7` is the pandas/time-zone database notation corresponding to UTC−07:00. During the study period, Hermosillo, Sonora used UTC−07:00 and did not observe daylight saving time.

Integrity verification from a Linux shell:

```bash
sha256sum mobility_hmo_v1_0_0.parquet
sha256sum raw_daily_csv_2020-09-18_2020-12-13.zip
sha256sum mobility_hmo_v1_0_0.csv
```

Compare the resulting hashes with `integrity/checksums_sha256.txt`.

## License

This dataset is made available under the Creative Commons Attribution 4.0 International License (CC BY 4.0). Users may share and adapt the material, provided that appropriate credit is given to the dataset creators, Universidad de Sonora is identified as the publisher, and LUMEX CONSULTORES, S.C. is acknowledged as the provider of the original mobility data source.