##### **TAID (Version 1.0)**

###### Welcome to TAID v1.0!



First off—a quick disclaimer: this project was built in isolation as a passion project by a solo author with a total budget of £50. Because of that, there may (is) some inefficient loops or repetitions in the code. These inefficient aspect also serve a purpose in memory retention. That being said I greatly welcome any feedback, performance optimizations, or future collaborations to help push this further.



If you run into any complications or just want to chat about the workflow, please feel free to reach out!



##### Workflow Overview

Due to the computational and memory demands of managing this workflow on standard hardware, the pipeline is currently split into four separate scripts to delineate distinct tasks:



1. 01\_cleaning\_and\_prep – Initial data cleaning and pre-processing.
2. 02\_cost\_layer\_MHWN – Generating cost layers for Mean High Water Neaps (MHWN).
3. 03\_cost\_layer\_MHWS – Generating cost layers for Mean High Water Springs (MHWS).
4. 04\_index\_creation\_and\_tiling – Final index creation and tiling of output products.



##### Configuration \& Prerequisites

Because local folder structures vary, you will need to manually set file paths in the scripts before running them. Please complete the following path definitions:



1. \[ \[ Insert variable / line location, e.g., Input raw data path ]
2. \[ ] 2) \[ Insert variable / line location, e.g., Output directory ]
3. \[ ] 3) \[ Insert variable / line location, e.g., Spatial reference / boundary layer ]
4. \[ ] 4) \[ Insert variable / line location, e.g., Scratch workspace ]
5. \[ ] 5) \[ Insert variable / line location, e.g., Intermediate raster storage ]



##### Hardware Recommendations \& Performance Notes

Strongly Recommended: Run all code and store temporary data on a fast, non-web-synced local SSD (avoiding active cloud syncing like OneDrive or Dropbox). In testing, this cut total processing time by roughly 50%.



##### Memory Management

You may spot repeated blocks of code—these are intentionally designed to clear memory as frequently as possible. While this trade-off slightly slows down total runtime, it prevents memory crashes and enables the processing of significantly larger spatial areas on standard desktop setups.



##### Benchmarks

The full 4-step pipeline takes approximately 2 hours to complete (with Steps 1, 2, and 3 taking up the bulk of the processing time).



Tested System Specs:

CPU: 12-Core @ 1300 MHz

RAM: 16 GB

GPU: 2 GB



##### Citation \& Context

For full context on the methodology, please refer to the associated manuscript. As noted there, TAID is very much an initial release (v1.0), and community contributions to refine the code are warmly invited!