Overall Description

[Debarpan08]

As of my understanding, the file structure and their respective functions are:

process_snap_data.py :seems to be the core of the project, orchestrating the processing pipeline for SNAP data. It reads transects, transforms and crops datasets, upscales temporal resolution, interpolates values along transects, and calculates derived variables like average wind direction and speed.

rmse.py : calculates the Root Mean Square Error between two sets of points, useful for assessing the accuracy of model outputs.

plotfiles.py : allows interactive visualization of raster files, aiding in data inspection and validation.

snap_tools.py : contains helper functions for data comparison, manipulation, and visualization, facilitating various data analysis tasks.

DeeringAutoDownload.py : appears to handle the automation of downloading satellite imagery, with potential support for different types of satellite data.

pad_tif.py : is responsible for padding TIFF files with additional bands of zeros.

ndwi_labels.py and label_inputs.py seem to preprocess image and label data, potentially for machine learning applications.

gsw_monthly_labels.py extracts monthly surface water labels from the Global Surface Water Dataset.

data_preprocessing.py contains functions for preprocessing raster data, such as splitting multiband images into tiles and augmenting image tiles for training datasets.

[Debarpan08]

Is that right? @fwitmer

[fwitmer]

For this stage of the project, don't worry about the SNAP climate data. The focus is on automatic coastline extraction. So you can ignore the process_snap_data.py and snap_tools.py files.

[Debarpan08]

Also, @fwitmer , using a U-Net CNN can be useful after we integrate it into the existing project pipeline for image analysis or processing. Additionally, the flexibility and scalability of U-Net will allow for adaptation to various tasks and datasets within the project domain, which will take into consideration the dynamic changes of annual wave energy.