Table 1.

Summary of simple linear regression analysis in the country and district domains.

statistic→ indicator name↓	assumed correlation from hypothesis	median R2 (district domain)ac	R2 (country domain)d	no. of districts with valid data (94 total districts)	no. of districts matching assumed correlation	no. of districts with significant models matching assumed correlationb	no. of districts with significant models not matching assumed correlation
maximum duration of winter cold spells	positive	0.20	0.06	80	58	15	1
days in winter cold spells	positive	0.13	0.05	80	54	12	1
mild winter weather	negative	0.13	0.03	80	56	5	1
maximum duration of late autumn cold spells	negative	0.10	0.01	78	50	4	4
days in late autumn cold spells	negative	0.10	0.00	78	47	2	4
days with optimal flying conditions	negative	0.07	0.02	80	62	4	1

^athe median is calculated from R²-values of districts with valid data that fit the assumed correlation.

^bonly models that fit the assumed correlation are counted.

^csee figure 4.

^dsee figure 3.