Fix the meteorology bugs identified in issue (#571)

# CABLE

Thank you for submitting a pull request to the CABLE Project.

## Description

Identified some meteorology bugs that caused issues during the later
spinup stages. Specific bugs are:

* Incorrect year was retrieved for previous/next day met variables
during spin-up
* Calculation of length of year became incorrect over time due to
implicit ```SAVE```

Fixes #[570](#570)

## Type of change

- [x] Bug fix

## Testing

This is not being pushed into the ```main``` branch.

Author: Whyborn

offline/cable_cru_TRENDY.F90

@@ -615,7 +615,9 @@ SUBROUTINE get_met_date(SimYear, SimDay, IsRecycled, RecycleStart,&
 
   ! We need to know how many days we expect in the year so check for end of year
   ! behaviour
-  INTEGER :: DaysInYear = 365
+  INTEGER :: DaysInYear
+
+  DaysInYear = 365
 
   ! Set up the first pass at recycling
   MetYear = SimYear
@@ -639,15 +641,15 @@ SUBROUTINE get_met_date(SimYear, SimDay, IsRecycled, RecycleStart,&
   ELSEIF (SimDay < 1) THEN
     ! Go back to last year- set the day later, once we know whether the MetYear
     ! is a leapyear or not
-    MetYear = SimYear - 1
+    MetYear = MetYear - 1
 
     ! This handles any future scenarios where we may change the day by >1
     MetDay = 365 + SimDay
     IF ((LeapYears) .AND. is_leapyear(MetYear)) THEN
       MetDay = 366 + SimDay
     END IF
   ELSEIF (SimDay > DaysInYear) THEN
-    MetYear = SimYear + 1
+    MetYear = MetYear + 1
     MetDay = SimDay - DaysInYear
   ELSE
     MetDay = SimDay