Differences between SWAT98.1 and SWAT99.2
prepared 4/2000 by S.L. Neitsch
updated 6/2000 by S.L. Neitsch
SWAT website: www.brc.tamus.edu/swat/

***************************************************************************** 
The final version of SWAT99.2 will be released by the end of May and will
include additional changes to the nutrient cycling in the soil as well as 
modifications to the pond/reservoir nutrient cycling. The version of SWAT99.2 
released in May will be the official SWAT99.2 code and will be the version of 
the code released on Basins 3.0. 
*****************************************************************************

Inputs
Input differences between SWAT98.1 and SWAT99.2 include: 
1) The full 4-digit year is now specified in all input files. This has affected 
the .pcp, .tmp and measured data (reccnst, recyear, recmon, recday) input files 
as well as the variable IYR in the .cod file.
2) creation of a new database, urban.dat (see Chapter 13 in new manual), that 
contains parameters related to urban land uses
Landuse maps used in the ArcView interface may be linked to the SWAT plant 
growth database (crop.dat) or the urban database (urban.dat) using the 4-
character code from either file. Obviously, if you add another land cover or 
urban land type, the 4-character code you assign to it must not already be 
present in either of these two files.
3) 2 urban land use variables, IURBAN and URBLU, added to the .sub input file
4) removal of all urban variables from the .mgt file except the street sweeping 
operation
5) street sweeping operation reassigned to management operation #12 (was #13)
6) new variables PERCOP and IRTPEST added to .bsn input file
7) a change in the definition of TMPSIM (formerly MSIM) and PCPSIM (formerly 
NSIM) in the .cod file. There are four options in SWAT98.1 and only two options 
in SWAT99.2
8) a change in units for SOL_ORGN, SOL_ORGP, and SOL_LABP in the .chm from kg/ha 
to mg/kg (ppm)
9) variables EPCO and ESCO moved from .bsn to .sub (can be defined for every HRU 
rather than for the entire watershed)
10) IDC in crop.dat is an integer instead of a real

Theory
Theoretical differences between SWAT98.1 and SWAT99.2 include:
Changes in water cycle:
1) lateral flow calculation altered:
C       COMPUTE LATERAL FLOW USING HILLSLOPE STORAGE METHOD
        if (j1.eq.1) then
          yy = 0.
        else
          yy = sol_z(j1-1,j)
        end if
        dg = sol_z(j1,j) - yy
        ho = 2. * adjf * sup / (sol_por(j1,j) - sol_fc(j1,j) / dg - 
     *      sol_wpmm(j1,j) / dg)
        if (ho.le.dg) then
c         ho = dg
          prk = ho * sol_k(j1,j) * slope(j) / slsoil(j) * .024
        else
          xls = slsoil(j) * (ho-dg) / ho
          prk = dg * sol_k(j1,j) * slope(j) / slsoil(j) * .024
          fup(j1,j) = xls * sol_k(j1,j) / slsoil(j) * 24.
        end if
to:
C       COMPUTE LATERAL FLOW USING HILLSLOPE STORAGE METHOD
        if (j1.eq.1) then
          yy = 0.
        else
          yy = sol_z(j1-1,j)
        end if
        dg = sol_z(j1,j) - yy
        ho = 2. * adjf * sup / (sol_por(j1,j) - sol_fc(j1,j) / dg)
!       ho = 2. * adjf * sup / (sol_por(j1,j) - sol_fc(j1,j) / dg - 
!    *      sol_wpmm(j1,j) / dg)
!       if (ho.le.dg) then
c         ho = dg
          prk = ho * sol_k(j1,j) * slope(j) / slsoil(j) * .024
!       else
!         xls = slsoil(j) * (ho-dg) / ho
!         prk = dg * sol_k(j1,j) * slope(j) / slsoil(j) * .024
!         fup(j1,j) = xls * sol_k(j1,j) / slsoil(j) * 24.
!       end if
3) checks inserted to make sure water content of soil layers does not go below 
zero when water is moved out of soil layer in any process
4) tile drainage flow added to model
Jeff has made a number of corrections/improvements to the nutrient cycling 
subroutines
1) residue decomposition and mineralization limited to top 2 layers. The minimum 
amount decomposed on any one day was lowered from .0075 * soil residue (98.1) to 
.0001 * soil_residue (99.2)
2) plant N and P demand takes into account root weight (excluded in 98.1)
3) plant P demand on any given day is increased by 50% to account for luxury P 
uptake
4) plant uptake of nitrogen and phosphorus altered
do 10 l = 1, nn
   unmx = uno3d * (1. - exp(-ubn * z(l,j) / z(nn,j))) / uobn
   unmx = amin1(unmx,uno3d-uno3)
   uno3l = amin1(unmx,wno3(l,j))
   uno3 = uno3 + uno3l 
now
do 10 l = 1, nn
   unmx = uno3d * (1. - exp(-ubn * z(l,j) / z(nn,j))) / uobn
   uno3l = amin1(unmx - uno3, wno3(l,j))
   uno3 = uno3 + uno3l
the same change was made in P uptake
5) nitrate percolation modified and soluble P percolation added
6) nitrate flux from second to first layer as a result of soil water evaporation 
processes added
7) fertilizer and auto-fertilizer application-organic N and P content was going 
into fresh organic N or P pool only in SWAT98.1; now split between fresh organic 
pool (50%) and active organic pool (50%)
8) nutrient stress calculation modified so that if amount of nutrient removed 
from the soil is 50% or more of the demand for that day, there is no nutrient 
stress for the day
9) in all nutrient calculations, qd (amount of surface runoff loading to main 
channel for day) is replaced with surfq (amount of surface runoff loading 
generated on day). qd incorporates lagging of surface runoff and is not the best 
choice for determining amount of nutrient removed
10) checks incorporated to ensure that when transformations between nutrient 
pools or removal of nutrients is performed, that the value of nutrient in a pool 
never assumes a negative value.
Other:
1) metals now routed through reach (only way to introduce metals to stream is 
through the measured data (reccnst, recday, recmon, recyear) files listed in the 
watershed configuration file

In June 12 release:
1) bank storage in main channel added to model
2) nutrient mineralization will occur only if the temperature is greater than 0
   degrees C.
3) added nutrient processes to pond/wetland/reservoirs. Reservoirs now route
   nutrients through.
4) variables added to .pnd and .res file