Annual ozone and moisture changes gave no clear repeated injury interaction in six US forest samples in 1994-2010

We tested whether annual ozone and soil-moisture changes jointly predict visible leaf injury after separation from site means. We reused 7,036 site-years at 1,276 US forest sites from a verified campaign publication. The pooled annual-change interaction was 4.29 percentage points per 10 ppm-h ozone times one moisture standard deviation, with a 95% interval of 0.11 to 8.46. All six region-period intervals included zero. Region-by-year controls gave 0.99 points, with an interval of -2.78 to 4.76. A joint test did not reject equal coefficients for the annual and site-mean products in the full sample (p = 0.432). Thus we did not establish that site differences explain the prior pooled result, or that an annual interaction repeats across samples. The package gives an exact product decomposition, code, aggregate results, source hashes and an independent numerical check. Exposure error, unknown missed injury, noisy site means and shared annual shocks limit inference. This is a reanalysis of the same source records, not a new independent archive.

ozone · soil moisture · foliar injury · within-site interaction · observation bias · forest monitoring · null replication