investigate correlation between readings at different weather stations

The correlations between the temperature readings is very high.

```
temp <- whately_2015 %>%
inner_join(orchard_2015, by = c("when" = "when"))
cor(Temp_C_Avg.x ~ Temp_C_Avg.y, data = temp)
[1] 0.9608951
```

Once we solve #9 and #8 it will probably be even higher.

But it also appears that the Whately station has higher readings for higher temperatures. Why? Is it because it's closer to the Sun?

```
xyplot(Temp_C_Avg.x ~ Temp_C_Avg.y, data = temp)
ladd(panel.abline(0,1))
```

Is this a real effect? Can we model it?

Start by reproducing the `inner_join` of `whatley_2015` and `orchard_2015`, then inspect the `cor` and `xyplot` results for `Temp_C_Avg.x` and `Temp_C_Avg.y`. Review issues #8 and #9 first; done means determining whether the station difference is real and documenting whether it can be modeled.

Written by the indexing model from the issue text.