LIS repeatedly reading LDT gen-metforcing at each 1h timestep

[bhanu-magotra]

I have a local metforcing dataset which is made to resemble a LDT generated metforcing. The dataset is arranged in the form of daily files containing multiple forcing variables. The files are named as following naming scheme :
LDT_HIST_198001010000.d01.nc
LDT_HIST_198001020000.d01.nc
LDT_HIST_198001030000.d01.nc
.... and so on.
My secondary dataset is MERRA2 which is also arranged in the form of daily files containing 24 timesteps each.
While running the LIS at 15 minutes timestep, I can check that for each day, MERRA2 is read-in only once but LDT gen-metforc is read-in at every step of 15 mins.

[INFO] LIS cycle time: 01/01/1987 00:30:00
 [INFO] Reading MERRA-2 file (bookend,           2  ... 
 ../../MET_FORCING/MERRA2/MERRA2_100/Y1987/M01/MERRA2_100.tavg1_2d_slv_Nx.198701
 02.nc4
 [INFO] Reading MERRA-2 file (bookend,           2  ... 
 ../../MET_FORCING/MERRA2/MERRA2_100/Y1987/M01/MERRA2_100.tavg1_2d_flx_Nx.198701
 02.nc4 (for lowest model level fields)
 [INFO] Reading MERRA-2 file (bookend,           2  ... 
 ../../MET_FORCING/MERRA2/MERRA2_100/Y1987/M01/MERRA2_100.tavg1_2d_flx_Nx.198701
 02.nc4 (for corrected precipitation fields)
 [INFO] Reading MERRA-2 file (bookend,           2  ... 
 ../../MET_FORCING/MERRA2/MERRA2_100/Y1987/M01/MERRA2_100.tavg1_2d_rad_Nx.198701
 02.nc4
 [INFO] Reading in forcing file: 
 ../../MET_FORCING/IMDAA_LDT_DAILY/FORCING/198701/LDT_HIST_198701020000.d01.nc
  - Reading and spatially reprojecting forcing fields - 
  Done reading and spatially reprojecting forcing fields - 
[INFO] LIS cycle time: 01/01/1987 00:45:00
 [INFO] Reading in forcing file: 
 ../../MET_FORCING/IMDAA_LDT_DAILY/FORCING/198701/LDT_HIST_198701020000.d01.nc
  - Reading and spatially reprojecting forcing fields - 
  Done reading and spatially reprojecting forcing fields - 
[INFO] LIS cycle time: 01/01/1987 01:00:00
 [INFO] Reading in forcing file: 
 ../../MET_FORCING/IMDAA_LDT_DAILY/FORCING/198701/LDT_HIST_198701020000.d01.nc
  - Reading and spatially reprojecting forcing fields - 
  Done reading and spatially reprojecting forcing fields - 
[INFO] LIS cycle time: 01/01/1987 01:15:00
 [INFO] Reading in forcing file: 
 ../../MET_FORCING/IMDAA_LDT_DAILY/FORCING/198701/LDT_HIST_198701020000.d01.nc
  - Reading and spatially reprojecting forcing fields - 
  Done reading and spatially reprojecting forcing fields - 
[INFO] LIS cycle time: 01/01/1987 01:30:00
 [INFO] Reading in forcing file: 
 ../../MET_FORCING/IMDAA_LDT_DAILY/FORCING/198701/LDT_HIST_198701020000.d01.nc
  - Reading and spatially reprojecting forcing fields - 
  Done reading and spatially reprojecting forcing fields - 
[INFO] LIS cycle time: 01/01/1987 01:45:00
 [INFO] Reading in forcing file: 
 ../../MET_FORCING/IMDAA_LDT_DAILY/FORCING/198701/LDT_HIST_198701020000.d01.nc
  - Reading and spatially reprojecting forcing fields - 
  Done reading and spatially reprojecting forcing fields - 

The effect of this turns out to be that with every passing year of simulation the run slows down massively. Eg, 1st year run took around 5hrs to complete, while the 3rd year run took 10hrs.
I can see in get_metForcGenerated.F90 that following code controls the read-in of 2nd forcing file but I need help in how to change it accordingly:

!- Determine if Forcing File 2 Needs to be Retrieved:
 ! LIS_time > Force_time1 (e.g., LIS_ts < Force_ts ):
   if( LIS_rc%time > metForcGen_struc%metforc_time1 ) then
      retrieve_file2 = .true.
    ! Assign Metforcing Data 2 to Metforcing Data 1:
      metForcGen_struc%metdata1(:,:,:)=metForcGen_struc%metdata2(:,:,:)

 ! LIS_time == Force_time1 (e.g., LIS_ts == Force_ts ):
   elseif( LIS_rc%time == metForcGen_struc%metforc_time1 ) then
     if( LIS_rc%ts == metForcGen_struc%ts ) then
       retrieve_file2 = .true.
     else
       retrieve_file2 = .false.
     endif
   endif

Here's ncdump of my local dataset for reference :

ncdump -h LDT_HIST_198001010000.d01.nc 
netcdf LDT_HIST_198001010000.d01 {
dimensions:
        time = 1 ;
        north_south = 476 ;
        east_west = 626 ;
variables:
        int64 time(time) ;
                time:time_increment = "86400" ;
                time:units = "days since 1980-01-01 00:00:00" ;
                time:calendar = "proleptic_gregorian" ;
        double lon(north_south, east_west) ;
                lon:_FillValue = NaN ;
        double lat(north_south, east_west) ;
                lat:_FillValue = NaN ;
                lat:units = "Degrees North" ;
        float Rainf_tavg(time, north_south, east_west) ;
                Rainf_tavg:_FillValue = NaNf ;
                Rainf_tavg:coordinates = "lon lat" ;
        float LWdown_tavg(time, north_south, east_west) ;
                LWdown_tavg:_FillValue = NaNf ;
                LWdown_tavg:coordinates = "lon lat" ;
        float SWdown_tavg(time, north_south, east_west) ;
                SWdown_tavg:_FillValue = NaNf ;
                SWdown_tavg:coordinates = "lon lat" ;
        float Psurf_tavg(time, north_south, east_west) ;
                Psurf_tavg:_FillValue = NaNf ;
                Psurf_tavg:coordinates = "lon lat" ;
        float Tair_tavg(time, north_south, east_west) ;
                Tair_tavg:_FillValue = NaNf ;
                Tair_tavg:coordinates = "lon lat" ;
        float Uwind_tavg(time, north_south, east_west) ;
                Uwind_tavg:_FillValue = NaNf ;
                Uwind_tavg:coordinates = "lon lat" ;
        float Vwind_tavg(time, north_south, east_west) ;
                Vwind_tavg:_FillValue = NaNf ;
                Vwind_tavg:coordinates = "lon lat" ;