Digitizing Historic Land Use Maps for Kern County

1984 historical land use map from Kern County

About

This repository contains scripts to digitize hand-drawn, georeferenced historical land use maps from Kern County, California. The maps are from the California Department of Water Resources (DWR). The ultimate goal is to identify lands that were once used for agriculture but have since been retired, and analyze how those areas may have revegetated over time. The digitization process involves several steps:

1. Georeferencing: Historical maps from 1958, 1969, 1977, and 1984 are georeferenced using Quad-G+ to align them with real-world coordinates
2. Digitization: The polygon_digitization.ipynb notebook processes the georeferenced maps to create polygons for each field and outputs each quad into a shapefile
3. Processing: The plotting.ipynb notebook combines individual quads into a single shapefile for analysis and visualization
4. Cleaning: The clean_up_testing/ folder contains initial explorations of polygon cleaning methods to remove noise and artifacts from digitized field boundaries

The excel_map_info folder contains Excel sheets with coordinate and metadata information for each map, which is used during both georeferencing and digitization to ensure spatial accuracy.

Data

This repository uses hand-drawn land use maps from the California Department of Water Resources (DWR) for the years 1958, 1969, 1977, and 1984. More information about these mapping efforts can be found in this ArcGIS StoryMap. The historical maps were scanned and georeferenced outside of this repository using Quad-G+. Once georeferenced, the maps are digitized in Python and the output is used for visual inspection and future analysis. Coordinate information for each map quad is stored in Excel files within the excel_map_info folder.

Repository Structure

historic_lulc_ag_lab/
├── polygon_digitization.ipynb # Main digitizing workflow
├── plotting.ipynb             # Plotting
├── excel_map_info/            # Coordinate and map metadata (excel files)
|   ├── quad_maps_1958.xlsx
|   ├── quad_maps_1969.xlsx
|   ├── quad_maps_1977.xlsx
|   └── quad_maps_1984.xlsx
├── clean_up_testing/          # Testing different methods for cleaning field polygons 
|   ├── comparison_test.ipynb  # Comparison between digitized maps and 1990 maps
|   ├── line_cleaning.ipynb    # Filter out diagonal lines
├── .gitignore                 # Files/folders excluded from Git
└── README.md                 

Georeferencing with Quad-G+

Before using this repository, the original hand-drawn maps must be georeferenced using Quad-G+. Both the application and detailed documentation can be found at this Box folder.

Setup Requirements

To use Quad-G+, you need CSV files containing information about each individual quad. These CSV files are located in the excel_map_info folder of this repository. Below is an example of the required file structure:

Quad-G+ Configuration

Preferences Settings

Use the following settings under the Preferences tab:

Output Setup

Customize the Output Setup tab to your desired file locations. The settings used for this project are shown below:

Important settings

• Set GCP Text Info (bottom right) to CSV
• Mode can be set to either automatic or manual
• For this project, automatic mode was used first, with manual troubleshooting as needed

Running the Georeferencing Process

1. Upload the CSV file: File > Open Batch File
2. Start georeferencing by clicking:
   • "Run Batch" (automatic mode)
   • "Find Corners" (manual mode)
3. Output GeoTIFF files will be saved to the folder specified in Output Setup

Usage

1. Ensure historical maps are georeferenced using Quad-G+ (see above)
2. Run polygon_digitization.ipynb to digitize the georeferenced maps
3. Use plotting.ipynb to visualize and combine individual quad shapefiles into a single shapefile
4. Reference Excel files in excel_map_info/ for coordinate and metadata information on maps

Polygon Cleaning

The digitized field polygons contain noise and artifacts from the original hand-drawn maps that need to be filtered out. The clean_up_testing/ folder contains preliminary explorations of different cleaning methods:

Current Cleaning Methods

Comparison Testing (comparison_test.ipynb)

• Compares digitized historical maps against an already digitized 1990 reference map
• Identifies matching and non-matching line segments between datasets
• Status: Preliminary exploration - requires further development

Line Filtering (line_cleaning.ipynb)

• Filters out diagonal lines that are likely map artifacts rather than field boundaries
• Status: Preliminary exploration - requires further development

Future Development

Both cleaning methods are in early stages and require significant improvement. Other cleaning methods should also be explored.

Acknowledgements

Thank you to Dr. Ashley Larsen for bringing me onto this project.