DAIRY HEAT STRESS preliminary web edition
Objective
How heat stress shapes U.S. dairy production
A report-aligned view of yield and milk-fat effects estimated with causal machine learning, together with realized national losses, state exposure, and year-specific evidence from 2003–2025.
Preferred cumulative effects per stress day
How to read this dashboard. Model estimates are pooled causal effects from the 21-state monthly estimation panel. Economic-loss views evaluate those pooled coefficients against realized exposure, production, and prices across the 48 contiguous states; they are not state-specific causal coefficients. Uncertainty ranges are coefficient-sensitivity bands.
National effects and realized losses
Compare the six estimators and trace annual loss sensitivity. Yield effects use approximate percent units; fat-content effects use percentage points.
Estimator comparison
Annual loss and sensitivity range
State explorer
Explore where realized stress, production, and prices translate into the largest cumulative or per-cow losses.
Cumulative heat loss
LowerHigher
State ranking
| Rank | State | Value |
|---|
Selected-state annual profile
Climate and production trends
Observed annual summaries for the 21-state estimation panel, with optional state detail. Missing fat-content months remain missing rather than being imputed.
Annual 21-state panel trend
Year-specific cumulative effects
Preferred NNET ADML estimates with 95% analytic intervals. These year-specific series are descriptive evidence of temporal heterogeneity.
Year-specific heat effect on milk yield
Sources, scope, and definitions
This file is a read-only snapshot. It contains all displayed data and code, requires no internet connection, and does not refresh automatically.
Source manifest
| Metric | Organization | Coverage | Project file |
|---|