When Not to Use Machine Learning in Hydrology
I train LSTMs and I build stochastic simulators. Here is the honest case for when machine learning is the wrong tool for a water problem — from someone who uses both.
Read →Field Notes
Writing
Essays on method and craft — when machine learning is the wrong tool, choosing a distribution, finding free data, and automating the work that should be automated. Grounded in real papers and real code.
-
-
Automating Flood Studies in HEC-RAS: When the GUI Becomes the Bottleneck
Driving HEC-RAS from code with the Controller API — what to automate, what to keep in the interface, and a real Guadalupe River / Hurricane Harvey workflow.
Read → -
How to Select Distributions in Hydrology — and Where to Find Them
A practical guide to choosing a probability distribution for frequency analysis — floods, low flows, rainfall — and the software that actually fits them.
Read → -
Satellite Precipitation for Indian Hydrology: A Practical Introduction
Why satellite rainfall matters for India's sparse gauge network, the products worth knowing, a three-step IMERG workflow, and the biases you must correct before you trust it.
Read → -
Free Hydrology Data Without the Hassle: Where to Download and Start Working
A working catalogue of free streamflow, precipitation, reservoir, and terrain data — and the client libraries that fetch it without the click-through misery.
Read →