Table 1 Table of Regression.

From: Climate impacts on long-term silage maize yield in Germany

Variable

Month

Specification

Silage Maize Anomaly SE: (Standard)/(Driscoll-Kraay)/(Bootstrap)

Precipitation

July

Polynom (degree 1)

0.264***

(0.028)/(0.033)/(0.110)

July

Polynom (degree 2)

0.001

(0.0003)/(0.001)/(0.001)

July

Polynom (degree 3)

−0.00001**

(0.00000)/(0.00000)/(0.000)

Temperature

July

Polynom (degree 1)

−6.443***

(0.634)/(1.001)/(5.840)

July

Polynom (degree 2)

−4.050***

(0.305)/(0.291)/(3.664)

July

Polynom (degree 3)

0.703***

(0.078)/(0.104)/(0.976)

Soil Moisture Index

June

severe drought

10.622***

(2.196)/(2.880)/(7.150)

June

moderate drought

8.723***

(1.988)/(2.303)/(3.960)

June

dry

3.198*

(1.722)/(1.763)/(2.561)

June

wet

−6.155**

(2.203)/(2.462)/(4.311)

June

abundantly wet

−12.173***

(2.660)/(3.813)/(5.767)

June

severely wet

−52.091***

(3.618)/(5.850)/(21.034)

Soil Moisture Index

August

severe drought

−47.447***

(2.609)/(3.820)/(12.549)

August

moderate drought

−21.952***

(2.066)/(2.837)/(5.985)

August

dry

−8.200***

(1.771)/(2.495)/(2.716)

August

wet

0.656

(2.084)/(1.800)/(4.000)

August

abundantly wet

−3.447

(2.428)/(2.431)/(5.881)

August

severely wet

−10.703***

(3.548)/(3.755)/(10.706)

Constant

  

18.905***

(1.155)/(1.527)/(4.710)

In-sample:

R2: 0.389

Adj. R2: 0.387

Adj. R2 - full variation: 0.705

LOCV (10-folds, 20 repeats):

R2: 0.385

RMSE: 37.014

MAE: 28.288

LOCV (annual blocks):

R2: 0.083

RMSE: 39.145

MAE: 30.589

LOCV (state blocks):

R2: 0.378

RMSE: 37.637

MAE: 29.169

Observations

4,625

  
  1. Standard errors are derived from three configurations. The first is the standard parametric configuration, and the second is the Driscoll-Kraay standard error, which parametrically accounts for serial and cross-sectoral autocorrelation and heteroscedasticity. The third configuration is based on a bootstrap approach resampling the years in the sample. The smallest standard errors are reported by the standard configuration, and the largest standard errors are reported by the bootstrap configuration.
  2. Note: *p < 0.1; **p < 0.05; ***p < 0.01; based on Driscoll-Kraay Standard Errors.