What have we learned from continuous crop and forest evapotranspiration observations in the Central Sands?

Ammara Talib, Ankur R Desai (2019)
Feb 5, 2019

Conference: ‘UW Extension & Wisconsin Potato & Vegetable Growers Association (WPVGA) Grower Conference 2019’

Ways to measure ET:

  • Meteorological approximations: Thornthwaite, Priestly-Taylor, Penman-Montieth – Strong assumptions on surface properties, ET driven by radiation

  • Residual: Measure precipitation, soil moisture, runoff, back out ET – Subject to high uncertainty Lysimeters: Labor-intensive, small-area estimate

  • Remote sensing: Assumptions on links between electromagnetic radiation and ET, satellites don’t work with clouds

  • Process models: Captures all aspects but requires extensive, hard to measure parameters

  • Eddy covariance: Direct method, whole field, expensive and expertise required

  • Machine “Learning”: Site-specific, hard to extrapolate or interpret, but powerful for forecasting

Machine learning Remote sensing Forecasting Models
Ammara Talib
Ammara Talib
Postdoctoral Fellow, Yale School of Environment

My research interests include using machine learning and statistical modeling to predict and forecast water quality and quantity issues and diseases.