How to Avoid Unnecessary Computation

The next stage of the solution consists of subtracting a multiple of the first row from the lower rows, to make the first element in each of these lower rows equal to zero. But the only nonzero elements in the first column are in the first two rows, so this manipulation need extend no further than the second row. The values in the first row are multiplied by the first element in the second row, and the resultant values are subtracted from the second row to convert the first element in the second row to zero. But 

DIM y(nrow), dely(nrow), yp(nrow), incind(nrow) 

DIM land(nlat), oceancl(nlat), landcl(nlat), zatemp(nlat) 

OIM Iwflux(nlat), xlat(nlat), albedo(nlat), insol(nlat) 

DIM hcap(nlat), cxlat(nlat), latmp(nlat), aveza(nlat) 

---