* NEWS --- Changes in the Daisy program and libraries -*- outline -*-

This file contains an overview of changes between Daisy source
releases.  Please note that the version numbers for binary releases
are less accurate, for example a binary version '1.40' may contain
some of the changes listed here under '1.41'.  You can get the version
number of the binary with the command 'daisy -v', or 'cdaisy -v' for
the daisy DLL.

When you upgrade Daisy, it is recommended that you read at least the
headlines of the changes since the last version.  For more detailed
information on changes, read the 'ChangeLog' file from the source
release.

* 2.93

** Groundwater pipe now survives flooding.

The pipe groundwater model does not works for a toally satiated soil
profile.  To work around this, it will now push water to the surface
before the first node get satiated.  This approximates some of the
effects of a groundwater table that lies above the soil surface.  But
if you log the groundwater table, it will be shown somewhere below the
surface.

You will get messages aboud flooding beginning and ending when this
happens.  If this happens a lot in reality, you should suggest the
farmer to improve his drainage system, it is obviously inadequate.  If
it only happens in simulations, you should increase either anisotropy
or K_aquitard, depending on whether you want the excess water in the
pipes or not.

Anyway, the effect is that the simulations survives in many cases
where they didn't use to.

* 2.92

** New anisotropy horizon parameter.

Set this to a number larger than 1.0 to make water flow faster towards
drain pipes.  

* 2.91

** Daisy now only print the date on the screen when something happens.

To restore the old behaviour, add

  (print_time daily)

to your setup file.

You can use any condition, so e.g. 

  (print_time hourly)

and

  (print_time (weekly 2))

will print time each our or every second week, respectively.

* 2.90

** Updated dk-horizon.dai with new conductivity parameters.

* 2.89

** Denitrification parameters changed meaning.

The 'alpha' and 'water_factor' parameters now only applies to the slow
pools.  On the other hand, the 'alpha_fast' and 'water_factor_fast'
parameters are now by default identical to the non-fast versions, so
everything should work as usual.

** New 'redox_height' Denitrification parameter.

All NO3 below this depth in the soil will be immediately denitrified
chemically.  You can log this with the new converted_redox log
variable, which is also in the Denitrification log parameterization.

By default, no chemical denitrification occurs.

* 2.88

** New 'M_vGp' hudraulic model by Bųrgesen et.al..

Compared to a straight Mualem + van Genuchten, it has two extra
parameters, 'h_m' which indicated the transition point for macropores,
and 'f' which is a shape parameter for the conductivity function.

** Clarified documentation for the DSnew harvesting parameter.

** Slight speedup in logging.

* 2.87

** Denitrification can now be affected by fast OM pools.

For a definition of "fast OM pools", see entry 2.71.

This is controled by two new parameters to the Denitrification
submodel, alpha_fast and water_factor_fast. You can set them like
this:

    (defcolumn Andeby default
      "The B.And farm, Andeby, 2002."
      (Denitrification (alpha_fast 0.3)
                   (water_factor_fast (0.5 0.0) (0.8 0.1)))
      (Soil (horizons (-20 [cm] Ap) (-2.5 [m] C))
      (MaxRootingDepth 60.0 [cm]))
      (Groundwater deep))
    
    (output "Denitrification")

These adds to the denitrification just like the old alpha and
water_factor parameters, but the total denitrification is still
limited by the K parameter.

By default, CO2 from fast pools have no special effect on
denitrification.  

You can get detailed information about potential and actual
denitrification with the new "Denitrification" log parameterization.

** CO2_fast is no longer printed in "Carbon Balance".

* 2.86

** More information from the "Carbon Balance" log parameterization

When choosing the "Carbon Balance" log parameterization, you will get
a summary like the following printer to the screen as well as in
daisy.log.   

Applied fertilizer =  300 [kg C/ha/y]
         Residuals = 1194 [kg C/ha/y]
             Roots =  325 [kg C/ha/y]
-------------------------------------
     Total C input = 1819 [kg C/ha/y]

In general, it is now possible to make Daisy print such summaries for
log files generated by the 'table' log model.  Look at the
documentation for the new 'summary' parameter, and at the "Carbon Balance"
parameterization in 'log.dai' for an example.

To get rid of the summary, specify an empty summary parameter for the
log parameterization, like this:

(output ("Carbon Balance" (summary)))

** The "N Balance" log model print out summaries for NH4 and NO3.

Similar to above.

** The "Root Zone Water Balance" log model now prints a total water balance.

Similar to above.

** The "Root Zone Water Balance" log model now includes surface storage.

Needed to create a total water balance.

* 2.85

** Minor cleanups to logging.

* 2.84

** New log table time_columns parameter.

If you specify this, you will automatically get year, month, mday and
hour columns.  By default, time_columns will be true if you do not
specify any time columns in the entries list.  All the log
parameterizations in log.dai has been updated to take advantage of
this, which gained approximately 10% speed.

* 2.83

** Minor cleanups and speedups to logging.

** The Surface lake parameter might work with lr.

* 2.82

** Minor cleanups and speedups to logging.

** The Surface lake parameter might work again.

* 2.81

** No user visible changes.

* 2.80

** Fixed missing "when" in log.dai.

* 2.79

** Log files now contain a line with initial content.

This line display the initial content of all non-flux variables.  It
only affect log parameterizations that has any non-flux variables.

** Changed default for OrganicMatter input root parameter to 800 [kg C/ha/y]

This reflect simulation results made by Henrik Svendsen.

** New duplicate warning.

If you specify the same attribute twice in the same list, you will get
a warning.  Same if you define a parameterization twice, or try to
overwrite a build-in model.

* 2.78

** The old log file syntax is restored.

You can now again begin log file parameterizations with

  (entries (year (path time))
           (month (path time))
           (mday (path time))
           (hour (path time))

but you will get a warning to shift to the new syntax mentioned in the
entry for 2.73.

** More control over the 'hypres' hydraulic model.

By default, the top horizon will be initialized by 'hypres' as a
topsoil, and the rest as subsoil.  You can now control this by
specifying the topsoil parameter.

(defhorizon Ap default
  (hydraulic hypres (topsoil true)))

(defhorizon C default
  (hydraulic hypres (topsoil false)))

** Changed default value for K_aquitard to 1e-4 [cm/h].

This groundwater pipe parameter used to default to 1e-5 [cm/h].

* 2.77

** A bit more speedups and cleanups to logging.

Having lots of log files is no longer expensive,

** The 'flux' parameter for all select models.

Instead of writing (when hourly) to ensure flux variables are getting
accumulated between log file timesteps, write (flux true).  The old
syntax will contibue to work for some time, but generates a warning.

** New dk-horizon.dai parameterization file.

This file contains median values from the DJF "Profil Database".

** You can now use [cm/d] instead of [cm/h].

* 2.76

** Logging is much faster.

Generation of log files is now seven times faster than before.  Lots
of stuff was changed in order to achieve this speedup, so it is likely
than some features have broken.  Please let me know in that case.

** Removed the table1 log model.

The one place it was used in log.dai (the PMSW log parameterization)
has been converted, fetch a new file.

* 2.75

** Minor fix in organic matter initialization.

* 2.74

** New automatic aquitard horizon.

Drained soil will now automatically get an aquitard horizon, unless
you specify the discretization (zplus).  The aquitard horizon is a
third of the size of the aquitard itself.

This affect logging, since the log models by default log down to the
end of the last horizon.  I therefore suggest you always specify the
'to' parameter when using the 'pipe' groundwater model.

The aquitard has the following parameters

(defhorizon aquitard default
  "Aquitard horizon used by drained soil."
  (hydraulic hypres (K_sat *K_aquitard*))
  (clay 50)
  (silt 20)
  (sand 29.99)
  (humus 0.01)
  (dry_bulk_density 2.0 [g/cm^3]))

where *K_aquitard* is the value of that parameter in the pipe
groundwater model.  

If you define your own horizon named "aquitard", that will be used
instead of the above.  The only requirement is that the hydraulic
model used must accept a 'K_sat' parameter.  Most do.

* 2.73

** Log file syntax changed.

To write year, month, day of month and hour to the log file, instead
of 

  (entries (year (path time))
           (month (path time))
           (mday (path time))
           (hour (path time))

the syntax is now

  (entries (number (path time year))
           (number (path time month))
           (number (path time mday))
           (number (path time hour))

The log parameterizations in log.dai have been updated.

** The "Organic Matter" log model now log SOM3 as well.

* 2.72

** New 'C_per_N' horizon parameter.

The intent of this parameter is to represent the measured C/N ratio
of the soil in the horizon.  Use it like this:

(defhorizon Ap default
  ;; Insert other parameters here.
  (C_per_N 16.0))

Note that the effect of setting this parameter is entirely unproven,
so use it with great caution.

The idea is that the varlues in the old 'SOM_C_per_N' parameter now
represents the "goal" C/N for the SOM pools, but not necessarily the
initial value.  The idea is that the farmland always end up with a C/N
ratio of 11, but that newly cultivated land may have a different
ratio.  We model this by saying that C/N entering the pool will always
have the ratio given in 'SOM_C_per_N', but C/N leaves the pool with
the current ratio, initially very close to the value specified by
'C_per_N'.

Note that the initial C/N ratio will not be identical to the number
spcified by the 'C_per_N' parameter, as 'C_per_N' represent the
average of all initial pools, including the AOM and the SMB pools.
The SOM pools will get what is left after the AOM and SMB pools have
been initialized.  This number will usually be very close to 'C_per_N'
though, as the SOM pools tend to much larger than the other pools.

Setting 'C_per_N' instead of 'SOM_C_per_N' for a soil with a high
measured C/N ratio will give significantly less mineralization, as
add organic matter will no longer for forced to have the same high C/N
ratio as the original content.

Currently all SOM pools will have the same initial C/N ratio when you
specify the 'C_per_N' parameter, but we might chang this so fast pools
ar closer to the goals specified in 'SOM_C_per_N' than slow pools.

* 2.71

** New CO2_fast log variable in OrganicMatter.

It is logged in the "Carbon Balance" log model.

It tells the CO2 produced by fast pools, where fast mean it has a
turnover rate larger than CO2_threshold, a new OrganicMatter
parameter.  The default value of 1e-4 means the SMB2 pool and all AOM
pools are considered fast with the default parameterization.

* 2.70

** New turnover_factor horizon parameter.

This parameter affects the turnover of all organic matter pools, in
effect making th whole system go slower (or faster if you set it
larger than 1).  Use it like this

(defhorizon C default
  ;; Insert other parameters here.
  (turnover_factor 0.1 []))

to make the organic matter system move 10 times as slow in the "C"
horizon. 

* 2.69

** Fixed bug that prevented B_C hydraulic model from being used.

* 2.68

** SOM1 now limited to be between 0.3 and 0.7.

If you let Daisy calculate the initial organic matter partitioning,
that is, do not specify the SOM_fractions horizon parameter, Daisy
will no longer come up with SOM1 fractions below 0.3 or above 0.7,
which is what our simulations have shown SOM1 stay within after change
in farming practice for typical Danish conditions.

These limits are controlled by three new OrganicMatter init
parameters, the default values are:

  (OrganicMatter (init (SOM_limit_lower 0.3 0.7 0)
                       (SOM_limit_upper 0.7 0.3 0)
                       (SOM_limit_where 0)))

Check the reference manual for a description of each.

** You can now specify K_sat or K_at_h for th hypres hydraulic model.

If you do so, that value will be used instead of the value predicted
by the HYPRES model.  All other hydraulic parameters will still be
calculated by HYPRES.

(defhorizon Ap default
  ;; Insert other parameters here.
  (hydraulic hypres (K_sat 3 [cm/h])))

** Allow automatic partitioning for subsoil with low humus content.

Previously, it would result in a negative SOM3 value.  Now we
repartition it as a top_soil, which mean SOM3 will be zero and SOM1
will not be in equilibrium.

* 2.67

** New K_at_h hydraulic parameter in most models.

This is an alternative way of specifying K_sat.

  (K_at_h 2 [pF] 5e-8 [m/s])

will mean the hydraulic conductivity have been measured to 5e-8 m/s at
pF 2.  From this, Daisy will find K_sat automatically.

** Changed EpFac for the simple and default crop models to 1.15.

Expect significant changes in simulation results.

You can change it back to 1.0 with

    (sow "Spring Barley" (Canopy (EpFac 1.0)))

in order to reproduce older results.

** 'soil_h' state variable removed from phenology models.
** 'soil_temperature' and 'partial_soil_temperature' state variables moved.

From phenology models to root_system submodel.  This will not affect
most people.

* 2.66

** The `weekly' condition should work now.

It never worked before.

* 2.65

** You can now initialize organic matter based on background miniralization.

If you specify 'background_mineralization' in OrganicMatter init and
do not specify 'SOM_fractions', the system will attempt to choose SOM1
and SOM2 values so the total mineralization from the SOM (and SMB)
pools, but not the AOM pools, are that amount.

Example

  (OrganicMatter (init (input 1400 [kg C/ha/y])
                       (background_mineralization 30 [kg N/ha/y])))

The net mineralization is the background mineralization plus the
amount of added organic nitrogen.

* 2.64

** There is now an inert SOM3 pool by default.

You can use this to "park" part of the humus so it will not contribute
to the mineralization by specifying a third number to the
SOM_fractions horizon parameter.  E.g if you instead of 

  (SOM_fractions 0.6 0.4)

write 

  (SOM_fractions 0.3 0.2 0.5)

half the humus in the system will be inactive, cutting 50% off the net
mineralization.

If you do not specify the third number, it will be assumed to be zero.
In effect, the current behavior (with no SOM3 pool) will be
maintained.

** Default organic matter initialization of subsoil changed.

If you do not specify SOM_fractions for a subsoil horizon, Daisy will
automatically park sufficient humus in the inactive SOM3 pool to get a
zero net mineralization.

** The AOM pools are now considered part of the humus for initialization.

The effect of this is that the SMB and SOM pools will be
correspondingly smaller, and you will get a slightly smaller
mineralization.

* 2.63

** New PenClayFac parameter for root penetration.

Example:

    (sow "Spring Barley"
         (Root (PenClayFac (0.0 0.8) (0.1 1.0) (0.3 1.1))))

This mean root penetration will be 20% slower with zero clay content
and 10% faster with 30% clay in the soil.  The default root
penetration speed is valid for soil with 10% clay.

** Use average daily temperature for root penetration.

We used to to use the temperature at midnight.  This fix might affect
simulation results slightly.

* 2.62

** Bug fixes and cleanups.

** svat PMSW now obeys the screen height climate parameter.

This means it is useful for tall vegetation, like forests.

* 2.61

** No user visisble changes.

* 2.60

** No user visisble changes.

* 2.59

** Bug fixes and cleanups.

** Daisy will now refuse to run unless it can open an daisy.log file.

* 2.58

** No user visible changes.

* 2.57

** Bug fixes and cleanups.

* 2.56

** Changes to reference manual.

Simple default values (like numbers) are now printed in "sample"
sections.

** More parameter checks.

** daisy.exe is fast again

For mysterious reasons, the new compiler started generating very slow
code, even slower than pre-2.47.  Switching to another computer seem
to have solved that, so Daisy is up to the speed of 2.47.

* 2.55

** Removed the obsolete NH4_evaporation fertilizer parameter.

Use 'volatilization' instead.  

** Cleanups.

* 2.54

** No user visible changes.

* 2.53

** You can now specify the minimum concentration for nitrogen uptake.

Example:

(defcrop "MyGrass" "Grass"
  "Don't assume zero sink for nitrogen uptake."
  (CrpN (NO3_root_min 1e-8 [g N/cm^3])
        (NH4_root_min 1e-9 [g N/cm^3])))

* 2.52

** No changes.

* 2.51

** Fixed bug in organic matter initialization.

* 2.50

** Macropores now enabled by default for clay soil.

If the combined humus and clay fraction of the top horizon is above
5%, macropores will be enabled by default.  The macropores will go
down to 150 cm, or the first non-clay horizon, or the position of any
drain pipes, which ever is higher.

You can disable macropores like done in this example:

(defcolumn Andeby default
  "No macropores"
  (SoilWater (macro none))   ; Disable macropores.
  (Soil (horizons (-20 [cm] Ap) (-2.5 [m] C))
        (MaxRootingDepth 60.0 [cm]))
  (Groundwater deep))

The macropores will be initiated when the presure go above -3 cm.

** Automatic generated discretion now limited by dispersivity.

No interval size will be more than twice the dispersivity.

** Checkpoints work again.

** More robust error handling.

* 2.49

** Fixed bug introduced by 2.48 that prevented use of organic fertilizer.

* 2.48

** SMB initialization improved.

The SMB pools are still initialized assuming equilibrium with the rest
of the system, but the method is now exact rather than an
approximation.  This will cause small changes in mineralization, less
if you are using a warming up period.

** Alternative SOM initialization method.

If you do not specify the 'SOM_fractions' horizon parameter, the SOM
pools will instead be initialized from the total yearly C input, with
an assumption that SOM2 will be in equilibrium with the rest of the
system.   You specify the C input with a new 'init' parameter to
organic matter, like this:

  (OrganicMatter (init (input 1400 [kg C/ha/y])))

If you specify this parameter, it will also be used for initializing
the SMB pools.  If you do not specify the 'init' 'input' parameter, a
default value will be calculated from the current input rate of the
initial AOM content specified by the 'am' parameter to
'OrganicMatter'.  The default value for the 'am' parameter is

  (OrganicMatter (am (root (weight 1.2 [Mg DM/ha]))))

corresponding to 1.2 tons of dead roots.  This value not likely to
produce any useful numbers whatsoever.

You can estimate the yearly input by creating a likely crop rotation
for the rpevious decades, and run it for 10 years with

  (output ("Carbon Balance" (when yearly)))

add the "Fertilizer_C", "Residuals_C_top" and "Residuals_C_soil"
columns, and take the average yearly value.

Having done that, you will likely want to initialize the 'root'
parameter to 'init' as well.  The 'root' should be the average
"Residuals_C_soil".  A full specification for the 'init' parameter
would be:

  (OrganicMatter (init (input 1400 [kg C/ha/y]) ;Default: use am input.
                       (end -20 [cm])           ;Default: end of first horizon.
                       (fractions 0 1 0 [])
                       (root 480 [kg C/ha/y])
                       (dist 7 [cm])
                       (T 7.8 [dg C])           ;Default: from weather file.
                       (h -100.0 [cm])))

The equilibrium depends on soil clay content, temperature and water
content.  The clay content will always be taken from the horizons.
The temperature will be taken from the yearly average specified in the
weather file, unless overwritten by the 'T' parameter above.  By
default, the water content is assumed to be at field capacity.

The input is assumed to come from two sources.

1) Old roots whose density will decrease exponentially with the depth
   as specified by 'dist', for every 'dist' cm you go deeper, the root
   density will be decrease to half, until it reach the
   'MaxRootingDepth' soil parameter.

2) (Non-root) crop residuals and organic fertilizer, which will be
   distributed uniformly into the first horizon, or if specified the
   value of the 'end' parameter.

Note that 'input' is the total input, including roots.  The default
values for the 'root' parameter corresponds to 1.2 Mg DM/ha/y.

The 'fractions' parameter determine where the AOM pools will end up,
the three numbers represent SMB1, SMB2 and SOM2 in that sequence.  By
default, everything will be eaten by SMB2 as shown in the example.

Most users will only want to set the 'input' and perhaps 'root'
parameters.

You can see the calculated initial values for the SOM and SMB pools in
the 'daisy.log' file, and follow their changes with the
"Organic Matter" log model.

If you specify a very high input for a soil with a very low humus
content, Daisy will be unable to use this method for initialization,
and stop with an error.  The amount of SOM2 where the system is in
balance will be higher than the total amount of C in the soil.  In the
real world the system will not be in short-term balance either, you
will see that as a quick build-up of humus in the soil.

In reality, most systems will not be in short-term balance, to reach
even a short-term balance you need to run with similar input levels
for decades which is unusual.  Therefore, you should set SOM_fractions
explicitly if you have any data that will allow you to estimate the
mineralization.

If you have measured nitrogen content at harvest, you can fit the
SOM_fractions parameter until the simulated nitrogen content matches
the measured nitrogen content.  Increase the SOM2 fraction if the
simulated nitrogen content is too low to speed up mineralization, and
increase the SOM1 fraction if the simulated nitrogen content is too
high to slow down mineralization.

* 2.47

** More parentheses required for fetilize action.

Before, you could write 

  (fertilize mineral (weight 80 [kg N/ha]))

Now you must write

  (fertilize (mineral (weight 80 [kg N/ha])))

This fixes an old bug, and makes it consistent with the syntax for
incorporation where the parentheses have always been necessary:

  (fertilize (mineral (weight 80 [kg N/ha]))
             (to -20 [cm]))

I apologize for the inconvenience.

** Made checkpoints work with hypres and Cosby_et_al hydraulic models.

** NH4 lost to volatilization  is now always counted in "fertilized_NH4".

Previously, it was only counted during surface application, not during
incorporation.

** Better handle end of weather data.

** Daisy can now be compiled with Intel C++ 7.0.

In my tests, it produces slower and larger code than GCC 3.2.

* 2.46

** Fix bug with van Genuchten hydraulic models (including hypres).

If you had multiple horizons using the hydraulic model, all used the
same value for M, affecting root uptake.

** Allow Daisy to run even with very steep hydraulic conductivity curves.

** Switched to from Borland C++ to GCC for building daisy.exe

This seems to cut nearly half of the simulation time, and the size of
the executable to less than half the old size.

Technically, the new version is created by cross-compiling from Linux
to MinGW32 with GCC 3.2, while the old version was compiled natively
with Borland C++ 5.0.1.

* 2.45

** New missing_years attribute for the default weather model.

This allows you to reuse the weather data for additional years.  For
example, if you have weather data for 1991 to 2000 in a file named
"weather.dlf", but want to run a simulation from 1987 to 2001, you can
specify the weather model like this.

(weather default "weather.dwf"
         (missing_years ((1987 1990) (1993 1996))
                        ((2001 2001) (1991 1991))))

This should be read as an order to Daisy to, in the period from 1987
to 1990, use the weather data from 1993 to 1996, and for 2001 use the
weather data from 1991.

This is particularly useful for running Daisy with a "warming up"
period where we don't have actual measurements.

In the example above we map two periods (1987-1990 and 2001), but you
can have as many periods as you like, so you can e.g. run a 100 year
simulation with just 10 years worth of climate data.

Note that Daisy will always use the actual data when available, so if
you in the example above have some data from 2001 but not all, Daisy
will not switch to 1991 data before it have used up the real data.

* 2.44

** New "ForcedLAI" vegetation attribute.

You can now force the CAI to have a specific value for part of the
year.  This is useful if you have measured the total CAI and want
Daisy to use those values for photosyntheses and transpiration (and
interception capacity).  Here is an example:

(defcolumn Andeby default
  "The B.And farm, Andeby, 2003."
  (Vegetation crops
              (ForcedLAI (1987 (100 2.0) (200 4.0))
                         (1988 (100 0.0) (150 1.5) (200 5.0))))
  (Soil (horizons (-20 [cm] Ap) (-2.5 [m] C))
        (MaxRootingDepth 60.0 [cm]))
  (Groundwater deep))

The code above should be read: In 1987 we start using forced CAI day
100 where it is 2.0 and end forced CAI in day 200 where it is 5.0.  We
use liniear interpolation between these two points, so in day 150 it
will be 4.0.  Before day 100 and after day 200 we use the simulated
CAI values.  In 1988 we start day 100 at CAI 0.0, increase linearly to
1.5 day 150, and from there again linearly to 5.0 day 200, where we
switch back to using simulated CAI.  All other years we use simulated
LAI.

Only paramterizations of the "default" crop model will be affected by
the "ForcedLAI" attribute.

If there are mulitple crops, the forced CAI will be distributed among
them corresponding to the relative size of their simulated CAI.  When
forced CAI is active, this is the only use of the simulated CAI.  

But the simulated CAI will be calculated anyway and can be logged.
Most log files will only log the CAI actually used which, when Forced
LAI is in effect, will not be the sum of Leaf, Stem and SOrg AI.  The
new "Forced LAI" log model will log both the used, forced and
simulated CAI.  By default, it will log the sum of all crops, set
$crop to log a particular crop alone as below:

(output "Forced LAI"
        ("Forced LAI"
         (set "$crop" "Spring Barley")
         (where "sbarley.dlf"))
        ("Forced LAI"
         (set "$crop" "Grass to grain")
         (where "grass.dlf")))

** Bug fix for permanent vegetation.

The litter was not put back to the system.

** You can now overwrite individual parameters in lists of submodels

For example, if you want to create a version pig_slurry (form
fertilizer.dai) with faster turnover, you can do it

  (defam my_slurry pig_slurry
    "Pig slurry with faster turnover rates."
    (om ((turnover_rate 5.0e-4))
        ((turnover_rate 3.0e-3)) 
        ()))

This specifies a slurry with three OM pools.  The pools have the same
parameters as pig_slurry, except that the first two pools have a
faster turnover rate.  Note that pig_slurry itself have three OM
pools, a slow pool, a fast pool, and a pool that go directly to SOM2.

If my_slurry had more OM pools than pig_slurry, you would have had to
specify all parameters for the additional OM pools.

The last empty set of parantheses in the definition of my_slurry is
needed, otherwise you would not have got a copy of the third pool from
pig_slurry. 

** Fixed assimilate partioning for several crop paramterizations.

Specifically "Winter Wheat" and varieties, "Pea" and "Ikuwala Maize".

This might cause some changes in yields, possibly apart from
"Ikuwala Maize" these changes should be insignificant.

** New checks and better descriptions for many parameters.

As these checks found errors in several crops, you need to update your
lib files.

* 2.43

** "Grass to grain" grass model now regrow after a full or late cut

I.e., this grass needs a tillage operation in order to kill it.  

WARNING: This also affects regrowth after a "normal" cut, ans the
paramterization is no longer backed by experimental observations.
I.e. it produce bogus results!  Use only when bogus results are better
than no results.  Otherwise, use the "Grass" parameterization.

** New TSum phenology model.

Simply specify a temeprature sum for each fase:: emergence, vegetative
and reproductive.

 (Devel TSum 
        (EmrTSum 100.0 [dg C d])
        (VegTSum 760.0 [dg C d])
        (RepTSum 740.0 [dg C d]))

* 2.42

** Fixed bug in hypres implementation.

K_sat was 24 times to large.

** irrigation_subsoil and irrigation_total

New log variables in the default bioclimate model.  The total
irrigation is now included in the "Root Zone Water Balance" log model.

* 2.41

** New "original" pseudo-model.

This keyword allows you to reuse the original parameters of an component
attribute, when creating a specialization. 

For example, if we have this definition of a foo parameterization:

  (deffoo default "old foo"
    (bar default (a 1) (b 2)))

and want a new parameterization that is identical, except "b" should
be 42, we had to write

  (deffoo "new foo" "old foo"
    (bar default (a 1) (b 42)))

i.e. specify the type and all attributes for "bar" once again.

Now we can change "b" alone

  (deffoo "new foo" "old foo"
    (bar original (b 42)))

Se next change for a more specific example.

** Change to default crop model.

The value of the "Devel" attribute in the "default" crop model should
now be a parameterization of the new "phenology" component.

In practice, this means "(Devel " should be replaced with
"(Devel default" in the crop parameterization files.  This has already
been done with the files in the daisy-lib.zip file.

For any derived parameterization where you only changed a few
parameters of the "Devel" submodel compared to the original, you can
use the new "original" keyword, like this

  (Devel original (EmrTSum  300))

to overwrite the EmrTSum parameter only.

** Total precipitation now in "Weather" log file.

This is just snow + rain.

** New "l" paramter in the "B_vG" and "M_vG" hydraulic models.

You can now specify the "l" parameter to van Genuchten.

** Bug fixes.

* 2.40

** You can now just specify "sand" instead of "fine_sand" and "coarse_sand"

In the horizon.

** New default hydraulic model: hypres.

This model replace "Cosby_et_at" as the default hydraulic model.  

"hypres" is a pedotransfer function for estimating parameters for the
van Genuchten/Mualem equation.  We believe it works better than
Cosby_at_al for European soil.

"hypres" requires you to specify the dry_bulk_density horizon
parameter, if you for some reason does not have that value, we suggest
you go back to the "Cosby_et_al" pedotransfer function, by adding 

    (hydraulic Cosby_at_al)

to your horizons.  Otherwise, add

    (dry_bulk_density 1.5 [g/cm^3])

to the horizon, replacing 1.5 with the measured value.

As usual, all the pedotransfer functions are unreliable, you should
use measured values for the hydraulic properties when at all possible.

** Don't require leaf to survive harvest, as stem may provide photosyntheses.

** Better messages at harvest.

* 2.39

** Changed default dispersivity to 5 from 6.

** New calibration of organic matter model.

Daisy now uses the numbers from Sander Bruun et al, "Calibration and
Validation of the Soil Organic Matter Dynamics of the Daisy Model with
Data from the Askrov long-term experiments", Soil Biology &
Biochemistry, 2003.

###########################################
# THIS MEANS OLD SETUP FILES ARE INVALID! #
###########################################

You need to readjust the 'SOM_fractions' horizon parameter in your
setup files.  For a system in perfect equilibrium (input of organic
matter equal tuinover), the value will be

  (SOM_fractions 0.49 0.51)

For farms with a long term humus build-up, more of the humus will be
in the second (fast) pool.  For farms where the amount of humus is
decreasing, slow first (slow) fraction will be larger.  New farmland
and farmland using mineral fertilizer tend to have decreasing humus,
and thus a larger fraction of humus in the slow pool.  Note that we
are talking changes over decades and centuries, if you switched to
organic fertilizer five years ago, that is unlikely to have a
measurable effect on the SOM partitioning.

For horizons below the root zone, if there is any significant amount
of humus may be leftover from old tree roots, and slowly decreasing.
In that case, the vast majority (maybe 95\%) is in the first, slow
pool.  

All added matter must be adjusted to the new parameters, the
fertlizer.dai file and the crop parameterizations in the daisy-lib.zip
file have been updated accordingly.

* 2.38

** You can now use "all" as crop argument to the crop_DM_over condition.

This will include the combined  weight of all crops on the field in
the test. 

* 2.37

** (OrganicMatter (clay_factor ...)) renamed.

It is now a component.  To get the old behaviour, use

(OrganicMatter (ClayOM old (factor ...)))

** New biomod clayom model.

This is still very experimental.

* 2.36

** Daisy can now handle dissolved organic matter.

You can have any number of DOM pools, they are specified as a "dom"
attribute of organic matter.  For each DOM pool you must add an extra
fraction to the end of the SOM and SMB fractions.  You _can_ add extra
fractions to AOM pools, in order to direct fertlizer directly to the
DOM pools.

The DOM pools are eaten by the SMB, and tranported with the water
unless adsorbed to the soil.

By default, Daisy has no DOM pools.

Here is an example of OrganicMatter with one pool:

          (OrganicMatter (smb ((C_per_N 6.7)
                               (efficiency 0.6 0.6)
                               (fractions 0 0.6 0 0.4 0.0)
                               (maintenance 7.5e-05)
                               (turnover_rate 7.708e-06))
                              ((C_per_N 6.7)
                               (efficiency 0.6 0.6)
                               (fractions 0 0.4 0 0.6 0.0)
                               (maintenance 0.000416667)
                               (turnover_rate 0.000416667)))
                         (som ((efficiency 0.4 0.4)
                               (fractions 1 0 0 0 0)
                               (turnover_rate 1.125e-07))
                              ((efficiency 0.5 0.5)
                               (fractions 0.9 0 0.1 0 0)
                               (turnover_rate 5.83333e-06)))
                         (dom ((C (M 0.0 [g/cm^3]))
                               (N (M 0.0 [g/cm^3]))
                               (adsorption none)
                               (diffusion_coefficient 4.6e-6 [cm^2/s])
                               (efficiency 0.0 0.4)
                               (fractions 1 0)
                               (turnover_rate 0 [h^-1]))))

And here is an example of an AOM pool that goes directly to DOM:

(defam C_N_high organic
  (om ((turnover_rate 1.0)
       (efficiency 1.0)
       (fractions 0 0 0 1)))
  (dry_matter_fraction 1.0)
  (total_N_fraction 0.1)
  (total_C_fraction 0.4))

*** New "DOM Pools", "DOM-C" and "DOM-N" log models.

The first give a summary of two DOM pools in the specified interval.
The two later give the carbon and nitrogen content in the DOM pools of
all layers.  You can get the content of a specific DOM pool with the
"$pool" option, otherwise you will get the content of all of them.

** You can now specify global radiation in [MJ/m^2/d] in the weather file.

** You can now specify speed in [m/s] instead of [cm/h].

** Accept fertilizer with only a single OM pool.

** Better error message when using undefined parameterizations.

* 2.35

** fertilized NH4 in log files now include NH4 lost to volatilization.

* 2.33

** More internal changes in organic matter... beware.

* 2.32

** Internal changes and bug fixes to organic matter.

Please report any problems.

** Better timestamp in daisy.log.

** Code cleanup.

* 2.31

** More information i daisy.log.

** Complex default values are now also written in the reference manual.

** Bug fix in the transport convection model.

** C API change:  daisy_printer_alist now takes four arguments.

* 2.30

** Assertion failures are now logged in daisy.log.

* 2.29

** Made 'height_start' and 'height_end' optional in default macropore model.

It will now use the last and first point of the 'distribution'
parameter as defaults.  Also, some consistency checks have been added.

* 2.28

** You can now specify heat_factor and water_factor for individual OM pools.

This allows you to specify that heat and water affect some types of
organic matter more than it affects other types.

Here is an example where heat factor has been specified for AOM1 and
water factor has been specified for AOM2

(defam "Andeby Pig Slurry" pig_slurry
  (om ((initial_fraction 0.72)
       (heat_factor (0.0 0.0) (20.0 1.0))
       (C_per_N 100)
       (turnover_rate 2.0e-4)
       (efficiency 0.60 0.60)
       (fractions 0.5 0.5 0.0))
      ((initial_fraction 0.18)
       (water_factor (-10000.0 0.0) (-100.0 1.0) (0.00 0.0))
       (turnover_rate 2.0e-3)
       (efficiency 0.60 0.60)
       (fractions 0.0 1.0 0.0))
      ((C_per_N 11)
       (turnover_rate 1.0)
       (efficiency 1.0)
       (fractions 0.0 0.0 1.0))))

You can also specify heat and water factor for the SOM and SMB pools.  

(defcolumn Andeby default
  (OrganicMatter 
   (smb ((C_per_N 6)           ; SMB1
         (heat_factor (0.0 0.0) (20.0 1.0))
         (water_factor (-10000.0 0.0) (-100.0 1.0) (0.00 0.0))
         (efficiency 0.6 0.6)
         (turnover_rate 7.708e-6)
         (maintenance 7.500e-5)
         (fractions 0 0.6 0 0.4 ))
        ((C_per_N 10)          ; SMB2
         (heat_factor (0.0 0.0) (20.0 1.0))
         (water_factor (-10000.0 0.0) (-100.0 1.0) (0.00 0.0))
         (efficiency 0.6 0.6)
         (maintenance 4.16666666667e-4)
         (turnover_rate  4.16666666667e-4)
         (fractions 0 0.6 0 0.4))))

And similarily for SOM.

** Dimension of canopy_washoff_coefficient chemical parameter has changed.

It is now a fraction, 0.0 means it stick to the canopy and is never
washed off, 1.0 means it is fully dissolved in the intercepted water.

You must update chemistry.dai and all pesticide parameterizations.

** New log variable `water_stress_days' in "Crop Production".

This gives an estimate of the number of days worth of production lost
due to water stress.  It is created by multiplying the water stress
for each hour with the fraction of the daily radiation received that
hour, and accumulating the result.

* 2.27

** Residuals are now logged in "N Balance" and "Carbon Balance".

** Allow macropores to go below the lowest node.

* 2.26

** Specifying Soil zplus is now optional.

If you do not specify it, Daisy will attempt to create a reasonable
default discretitation.

** New log parameterizations "Organic Matter" and "AOM Pools".

These log the content of the individual organic matter pools.
"Organic Matter" logs SOM, SMB and all the AOM's.  "AOM Pools" logs
all the AOM's, or a specific AOM if "$am" is set.

Example: 

  (output "Organic Matter"
          ("AOM Pools" (set "$am" "Grass/root")))

** New spec attribute in select entries of the log table model.

A "spec" is a suplement to "path" for specifying which daisy attribute
should be logged.  If you specify "spec", Daisy will be able to either
find the dimension itself, or if you specify the dimension, attempt to
convert from the dimension used internally.

Example:

  (interval (path column "$col" SoilWater Theta)
            (spec column default SoilWater Theta)
            (tag "Water")
            (dimension "mm"))

The advantage of doing it this way is that Daisy will be able to
detect any errors.

The feature is still experimental, and will only work in few cases.
You will get an error for the cases where Daisy cannot convertion the
dimensions.

* 2.25

** Changed the equilibrium tranformation model.

The 'A' and 'B' parameters that specify the names of the two chemicals
are no longer ordered.

** You can now specify 'turnover_halftime' in organic matter pools.

This is an alternative to 'turnover_rate'.

** Changes to the default chemical model.

*** decompose log increment for pesticided moved from solute to chemical.

So now write 

 (defchemical Foo default
    (lag_increment (0.0 1e-5) (1.0 1e-4))
    ...)

instead of 

  (defchemical Foo default
    (solute (lag_increment (0.0 1e-5) (1.0 1e-4))
            ...) ...)

*** canopy_dissipation_rate_coefficient made obsolete.

Instead, you the new name 'canopy_dissipation_rate' or the new
parameter 'canopy_dissipation_halftime'.

*** New decompose_halftime parameter

It can be used instead of 'decompose_rate'.

** New [d] to [h] convertion.

Particular useful in konjunktion with the new halftime parameters.

* 2.22

** Soil water conductivity is now temperature dependent.

The specified water conductivity (K_sat) is assumed to apply at 
20 dg C, and scaled according to the water viscosity at the current
soil temperature.

* 2.21

** Added cross references to reference manual.

In particular, add information about where each component, model,
parameterization and submodel is used.

* 2.20

** Change the way to specify chemicals for logging.

You can now use a (set "$chem" Atrazine) parameter for "Surface
Chemicals", "Soil Chemicals" or "Chemicals" to log Atrazine alone, or
leave it out to log the sum of all chemicals in the soil.  As part of
this cleanup, the "Chemical" and "Soil Chemical" parameterizations has
been moved to 'log-old.dai'.

Also, Daisy will now automatically delay the Chemicals log until the
chemical enters the soil.  This means the number of columns will be
correct from start.  

** Change the way to specify organic matter for logging.

Update the "Total Soil Content", "N Balance", AOM and Bioincorporation
log parameterizations.  For AOM, use like (set "$am" "grass/root") to
log a specific AOM.  Compared to before, the "?" is missing.

* 2.18

** Daisy now creates a daisy.log file.

It contains most of the information printed during the run.

* 2.17 

** Fixed many (small) memory leaks.

Also added workaround for a memory bug in Groundwater pipe, and fixed
some uninitialized variables in svat PMSW.

** More cleanup in log.dai.

As usual, look in log-old.dai for missing log parameterizations, and
remember to use an explicit "when" if you don't want hourly values.

** New warnings and errors on invalid input.

* 2.16

** New "Prod RtDR_T_factor" parameter to the default crop model.

Temperature dependent factor for root death rate.

Use it like this

(defcrop MyGrass Grass 
  (Prod (RtDR_T_factor (0.0 0.0) (10.0 1.0))))

The will stop root death when soil soil freeze, and let it gradually
grow to full power when the soil temeprature reached 10 dg C.

** New PrecipCorrect keyword for the default weather model.

Add a 

  PrecipCorrect: 1.12 1.10 1.00 1.20 1.40 1.60 1.12 1.10 1.00 1.20 1.40 1.60

to the keyword section of your .dwf file to get mothly correction for
the precipitation.  The Precip-Correct should be followed by 12
numbers, each a factor to multiply to the precipitation for that
month.  The first number corresponds to January, the last to December.

** Made "Soil Temperature" and "Surface Water Balance"log models hourly.

** Moved "Daily Water Flux" and "Hourly Water Flux" to 'log-old.dai'.

* 2.15

** Amount of solute in drainage was entirely wrong.  Fixed.

** Moved all V1 crop parameterizations to 'crop-old.dai'.

** New soil_water_content_above condition.

See the section about irrigation in the tutorial for an example.

** Updated the pig_slurry and cattle_slurry parameterizations. 

** New pig_manure, cattle_manure, and horse_manure parameterizations.

** Updated Rye, Sugar Beet, and Spring Rape parameterizations.

** Renamed Maize to Pioneer Maize, and added new Maize parameterization.

The new maize is a sort common in Denmark.

** New Pendimethalin pesticide.

** New Potato crop parameterization.

** The reference manual is improved.

It now list standard parameterizations, and index all component,
model, submodel, attribute and parameterization names.  The
introduction is also updated.

* 2.14

** Fixed bug introduced in previous version that disabled macropores.

** Made check for mass balance more robust.

* 2.12

** Fixed error with extra CH2O in the default crop model.

It should not have measurable effect on production, but could produce
warnings during simulation.

** Made the default mactrans model more numerically robust.

This also removed some rare warnings during simulation, but has no
effect on the results.

* 2.11

** Fixed mass balance error in mineralization.

When removing tiny organic matter pools, their remaining C was
incorporated into the SOM, but not their N.  This could add up to 1 kg
N inbalance over 10 years under some circumstances.

** Imobilization is now restricted by available nitrogen.

It used to run on full speed, even when there was no nitrogen to
immobilize.

** Fixed bug in the Cosby hydraulic model.

The estimated conductivity was far too low to be useful.

** The cd transport model can switch to smaller timesteps.

This will happen in the cases where it gave negative answers, and
eliminates the "BUG: M[0] = -1.4115e-7 after transport" messages.  The
effect of the bug would be a slightly too low estimate on leaching for
nitrogen starved systems.

* 2.09

** There is now an index of logable variables in the reference manual.

** The 'default' bioclimate model is now default.

I.e., you don't have to specify it explicitly.

** If you don't specify DSnew in the default crop model, it will use 
the stub length to calculate the DS after a cut. 

* 2.08

** Fix bug preventing macropores and drainage to be used together.

** Fix checkpoints.

** You can now specify [d^-1] instead of [h^-1].

** You can now specify fractions with [%].

* 2.07 

** New "Cosby_et_al" hydraulic model.

This is a modified Bruce Campell model with parameters derived by
pedotransfer functions as specified by Cosby et al.

This model is now default, so you don't have to specify hydraulic
parameters for a horizon (you should, though, the pedotranspher
functions aren't that reliable.)

* 2.06 

** You can now specify [m] or [pF] instead of [cm].

** Syntax for specifying dimensions in arrays changes.

Dimensions now applies only to numbers before them.  For example:

  (foo 10.0 20.0 [cm] 1.0 1.5 [m] 100.0)

Here the two first entries (10.0 and 20.0) are in cm,  the two next
(1.0 and 1.5) are m, and the last is whatever the default dimension
for foo is.

* 2.05 

** New 'linear' Pedotransfer model.

This allows you to to specify parameters as a linear function of
horizon attributes.  Here is an example:

(defpedotransfer "van der Zee & van Riemsdijk" linear
  (description "0.5 (Al_0 + Fe_O)")
  (factors ("Al_O" 0.5)
           ("Fe_O" 0.5))
  (offset 0.0))
           
(deftransform "Phosphor Fast<->Slow" equilibrium
  (A "Phosphor Fast") (B "Phosphor Slow")
  (equilibrium Langmuir
               (K 1e-5 [g/cm^3])
               (pedo_my_max "van der Zee & van Riemsdijk")
               )
  (k_AB 1e-4))

Currently only the 'K' and 'my_max' attributes to the Langmuir
equilibrium model, and the 'k_AB' and 'k_BA' attributes to the
equilibrium transform model can be specified with pedotransfer
functions. 

The "Al_O" and "Fe_O" attributes must be specified in the horizon,
like this:

(defhorizon top default
  (attributes ("Fe_O" 1e-5)
              ("Al_O" 5e-6)
  ...)

* 2.04

** Transformations.

You can now create arbitrary transformations between soil chemicals. 

Example:

(defcolumn test default
  (Transformations (equilibrium (A "Phosphor Solute") (B "Phosphor Solid")
                                (equilibrium Langmuir
                                             (K 1e-5 [g/cm^3])
                                             (my_max 1e-2 [g/cm^3]))
                                (k_AB 1e-2 [h^-1])))
  ...)

Right now, "equilibrium" is the only implemented transform model.  It
transforms towards an equilibrium, using k_AB (and k_BA) as speed.
k_BA are by default the same as k_AB.

Langmuir is the only implemented equilibrium model.  

All the numeric parameters (K, my_max, k_AB, k_BA) are arrays, so they
can differ for each soil interval.  There are "pedo_" versions
(pedo_K, etc.) of them all, making it possible to specify them using
pedotransfer functions.  No pedotransfer functions are currently
implemented. 

** New "full" adsorbtion model

Use it with "none" transport to get non-solute soil chemicals.

** New "solid" standard chemical parameterization.

It uses the full adsorbtion model together with the "none" transport. 

** New "Soil Chemical" log parameterization.

These allows you to log a single soil chemical.

** New "Transform" log parameterizations.

This allows you to log transformations.

** Bug fixes.

* 2.03

** You can now specify dimensions for numbers in input files.

The dimensions are in brackets, like this:

(defhorizon top default
  (hydraulic M_vG 
             (K_sat 19.3 [cm/h])
             (Theta_res 0.05 [cm^3 H2O/cm^3])
             (Theta_sat 0.424 [cm^3 H2O/cm^3])
             (alpha 0.069 [cm^-1])
             (n 1.527 []))
  (SOM_fractions 0.66 [] 0.34 [])
  (SOM_C_per_N 11 [g C/g N] 11))

You must write the dimension exactly as is specified in the manual,
with the convention that ^ is used for superscript.  Also, fractions
and dimensionless values are both denoted as [].  Some values in Daisy
have unknown or unbspecified dimensions in the manual.  For these,
Daisy will accept any dimension in the input files, as long as it
start with a question mark.

There are no conversions yet, only the dimension used in the manual
may be specified.  In a furture release, Daisy will be able to make
some simple conversions, like [m] to [cm], allowing you to use what is
most convenient in the input file.

For arrays, if you specify the dimension for an entry (like
SOM_C_per_N in the example), that dimension is assumed to apply to the
rest as well, until something 

* 2.02

** You can now check on temperature sums.

To sow spring barley after the temperature sum reaches 110 dg C d,
write:

    (wait (TSum_above 110))
    (sow SpringBarley)

TSum is the sum of the daily average air temperature, ignoring frosty
days.  By default, it is reset March 1 every year.  

* 2.01

** New root density models

*** DS_Depth

Example:

    (sow "Spring Barley"
         (Root (rootdens "DS_Depth" 
                         (entries (0.0 (0.0 1.0) (10.0 0.0))
                                  (1.0 (0.0 1.0) (50.0 0.0))
                                  (2.0 (0.0 1.0) (40.0 1.0) (50.0 0.0))))))

This specifies the root distribution at DS 0.0 (emergence), 1.0
(flowering) and 2.0 (ripe) respectively.  At all times, the highest
density is at the top.  At emergence, the density diminish linearly
until 10 cms depth, below that there is no root.  At flowering, it
still diminish linearly, now down to 50 cm.  At ripe, it has maximum
desnity down to 40 cm, at which point it diminish linearly downto 50
cm. 

Daisy will interpolate between these three development stages.

*** DS_Rel

Example:

    (sow "Spring Barley"
         (Root (rootdens "DS_Rel" 
                         (entries (0.0 (0.0 1.0) (1.0 0.0))
                                  (1.0 (0.0 1.0) (1.0 0.0))
                                  (2.0 (0.0 1.0) (0.8 1.0) (1.0 0.0))))))

This works like DS_Depth, except that instead of using absolute
numbers for soil depth, we now use numbers relative to the current
potential root depth, as calculated by the root submodel.  With
DS_Depth, the calculated root depth is ignored.

*** Depth_Depth

    (sow "Spring Barley"
         (Root (rootdens "Depth_Depth" 
                         (entries (00.0 (0.0 1.0) (10.0 0.0))
                                  (50.0 (0.0 1.0) (50.0 0.0))
                                  (75.0 (0.0 1.0) (40.0 1.0) (75.0 0.0))))))

This works like DS_Depth, except that instead of using the development
stage as index, we now use the calculated root depth.  Obviously, this
doesn't work when the root distribution change while the root depth
remain constant, as implied by the DS_Depth example.

In general, it is recommended that DS_Depth or Depth_Depth is used
when actual measurements has been made, and DS_Rel is used for general
xscrop parameterizations since soil conditions may prevent the root from
reaching the same depth as the measured values.

* 2.00

** GNU/Linux i386 port.

* 1.99

** Harvest cut_delay in the default crop model changed.

It is now a function of the amount of dry matter removed (measured in
kg/ha), rather than of the fraction of dry matter removed.

This affects none of the standard parameterizations.

** Harvest cut_delay now limits development as well as production.

** Crop height is now limited by WStem in the default crop model.

The Canopy parameter 'HvsWStem' put an upper limit on the crop height
based on WStem, relative to the crop height at flowering.  By default,
it need 200 g DM / m^2 to reach full height.

This affects stem and leaf harvest for monocrops, and light
competition for multicrops.

** Change LAI distribution in default crop parameterizations.

LAIDist0 and LAIDist1, which controls how the leafs are distributed
vertically, have been modified for all parameterizations of default
crop model to something that makes more sense.  

This will only affect leaf harvest for most simulations, but have
great effect for grass with multiple cuts, and for multicrops.

** You can now set MinDens in the Gerwitz+Page74 root density model.

** New "Root Density" log model.

* 1.98

** Better errors for (some) illegal parameter values.

You now get line and column number for where they appeared.  

Lots of code was changed to do this, so if some of the errors seem
wrong, let me know.

* 1.97

No user visible changes.

* 1.96

** New 'cut_stress' factor in the standard crop model.

After a cut production may be delayed for some days, depending on how
large a fraction that was removed.

(defcrop "Grass with Cut Delay" Grass 
  (Harvest (cut_delay (0.4 7) (0.8 28))))

** Bug fixes.

* 1.95

** Bug fixes.

* 1.94

** Log file change.  

Moved remaining CrpAux variables to Prod in the standard crop model.
The log.dai file has been updated.

* 1.93

** Log file change.  

StemRes was moved from CrpAux to Prod in the standard crop model.

* 1.92

** Log file change.  

Phenology has been merged into Devel in the default crop model.  This
mean log parameterizations logging DS must be updated.  

A number of log variables have been moved from CrpAux to CrpN in the
default crop model.  The most likely to be logged are Fixated and
AccFixated. 

The parameterizations in 'log.dai' has been updated.

** "N Balance" now logs fixation.

* 1.91

** Added description string for fertilizer parameterizations.

* 1.90

** Now give better error old style root AOM initialization.

** Deposition is now logable, and logged in "N Balance".

* 1.89

** Root density is now a component.

This means that the 'SpRtLength' and 'DensRtTip' root parameters have
changed, and that parameterizations that sets these must be updated.

Instead of 

(defcrop MyCrop default
  ;; ...
  (Root (SpRtLength 100.0)
        (DensRtTip 0.1)

use 

(defcrop MyCrop default
  ;; ...
  (Root (rootdens "Gerwitz+Page74"
                  (SpRtLength 100.0)
                  (DensRtTip 0.1))

If you don't have any better value for SpRtLength or DensRtTip than
the above, you don't need to specify them.  They are the default
values.
 
* 1.88

** "N Balance" and "Carbon Balance" now contains harvest information.

* 1.87

** Mark "CrpN SeedN" as obsolete in default crop model.

Use "Prod NCrop" instead.  You will need to update the crop library to
get rid of the warnings.

** Make it possible to log nitrogen added with seeds.

It has been added to the "Fertilizer" log parameterization.

** Lots of updates to the log parameterizations.

* 1.86

** Count initial AOM when estimating initial SMB.

** By default, assume 1.2 t DM/ha old root residuals at initialization.

To disable, specify an empty value for "am" in OrganicMatter.

The effect if the above two changes is a much reduced need for a
"warming period" at the start of the simulation. 

** The syntax for "root" AM has changed, everything has default values.

Typically, you would just override the "weight" variable with your own
estiimate.

* 1.85

** New top level attributes "directory" and "path".

You can now specify where Daisy shall put log files, and where it
shall look for input files, directly from the ".dai" file.  The
following should be put near the top of the dai file:

   (directory "C:/my/daisy/setup/output")
   (path "." "C:/my/daisy/setup" "C:/daisy/lib")

This will make Daisy run in the directory "/my/daisy/setup/output" (or
"C:\my\daisy\setup\output"), and also generate its log files there.

It will search for input files first in the directory where it run
("." means "current directory"), if it doesn't find them there it will
look in "C:\my\daisy\setup", and if it isn't there either, it will
finally look in "C:\daisy\lib" before giving up.

PS: Remember to use "/" instead of "\" for file names in ".dai" files.

** Bug fixes.

* 1.83

** More Borland work.

** Don't log C harvest per default.

* 1.82

The harvest log model now gives information about the carbon content.

* 1.80

** Now compiles with C++Builder 5.0 via the commandoline and Cygwin make

** Now includes a C++Builder 5.0 project file.

** Now compiles out of the box with Cygwin and "make daisy.exe"

** Bug fixes.

* 1.79

** Log the evapotranspiration used, not just the one in the weather file.

** Bug fix: ignored last root node in some circumstances.

* 1.78

** Keep track of C in crop.

Note: This means the CStem, CDead, CSOrg, CRoot and CLeaf harvest
parameters of the standard crop module can no longer be set.  Instead
set the E_Leaf, E_Stem, E_SOrg, E_Root production parameters.  You can
also log the C content with the CStem, CSOrg, CRoot and CLeaf in
production. 

* 1.77

** Ported to GCC 3.0

** You can now plant emerged crops.  

Define the crop like below, and use the normal "sow" command.

(defcrop "Maize for Planting" "Maize"
 (Phenology (DS 0.269081))
 (Prod (CH2OPool 12.6672)
       (NCrop 1.41396)
       (WLeaf 15.7116)
       (WRoot 26.4654)
       (WStem 10.9263))
 (Root (Depth 51.9679)))

* 1.76

Fix logging state for 'with-column'.

* 1.75

Fix logging of speed in Bioincorporation.

* 1.74

** More flexible dimensions in logs.

The character & in a dimension string in a log description will be
replaced by "h" if you use "hourly" (with no arguments) as a log
condition, "d" if you use "daily" (with no arguments) as a log
condition, "w" if you use "weekly" (with no arguments) as a log
condition, "m" if you use "monthly" (with no arguments) as a log
condition, "y" if you use "yearly" (with no arguments) as a log
condition, and "dt" otherwise.  All the standard log descriptions
have been updated to use "&".

This means that if you for example put

  (output ("N Balance" (when weekly)))

in your setup, "N Balance" will be logged weekly, and the flux
dimensions will be correct.

** Fixed off-by-one error for flux bottom in logs.

* 1.73

Fixed logging incorporated inorganic matter.

* 1.72

Fixed some dimension strings.

* 1.71

Fixed bug in PMSW checkpointing.

* 1.69 

Fixed bug in Penman-Monteith.

* 1.68

NH4 Volatilization is now in the Fertilizer log.

* 1.66

Disabled 'old' and 'old2' hydraulic models, and 'file' and 'hourly'
weather models.

* 1.60

** New 'EpFacDS' Canopy parameter.

This PLF allow the crops potential evapotranspiration adjustment to
depend on the development stage of the crop.

* 1.59

** Checkpoints should work now.

* 1.58

** Tillage can now affect background mineralization.

See the 'som_tillage_factor', 'smb_tillage_factor' and 'tillage_age'
parameters to 'OrganicMatter'.

* 1.55

** New 'density' flag to the 'interval' select method for logging.

Set 'density' to true to make Daisy log the average (in height)
content of the interval, rather than the total content.

** Fixed bad initialization crash.

Found by ML.

** The PM pet module now gives more realistic results.

Bugs found by SH.

* 1.54

** Adjusted DSLAI05 in 'Grass' parameterization.

Should give larger initial harvest.

** The ScreenHeight weather parameter is now respected.

This affects the 'PM' potential transpiration model.

* 1.51

** Checkpoints now include input statements and definitions automatically.

The 'libraries' attribute has been removed.
Checkpoints still doesn't work, though.

* 1.50

** Minor reference manual enhancements.

It will now state if submodel as a full or partial value.

* 1.48

** Fix 'monthly' and 'yearly' conditions.

Used to match every hour in the month or year, except the last.
Reversed that.

** You can now log applied fertilizer.

Use the new 'Fertilizer' log model, defined in 'log.dai'.

** Fix loading global files (bug introduced in 1.47).

* 1.47

** New default SMB and SOM parameters. 

These should have little effect on mineralization, but give a more
realistic estimate of the size of the SMB pools.

It *will* have a large effect if you are running without a starting
period, and have large initial AOM pools.

** New fertilizer types in 'fertilizer.dai'.

** More work on the winter wheat and winter barley parameterizations.

* 1.46

** Fixed link order dependency crash (MAYBE).

* 1.45

** Fixed crash related to the am/cleanup AOM introduced in 1.40.

** Defaults for the 'minimum', 'from' and 'to' precision fertilize params.

* 1.44

** New 'first_year_utilization' fertilizer parameter.

Use this together with the new 'equivalent_weight' fertilize
parameter, to let Daisy calculate the amount on organic fertilizer to
apply, to get an inorganic nitrogen equivalent.  Example:

  (fertilize (pig_slurry (first_year_utilization 0.20))
             (equivalent_weight 50.0)) ; kg N/ha

Here, we specify that 20% of the nitrogen (organic or inorganic)
contained in pig slurry can be utilized the first year, and that we
want to fertilize with an amount of pig slurry corresponding to 50 kg
N/ha.  Daisy will then apply, say, 3 ton DM/ha pig slurry, depending
on the N content of the slurry.

** New 'precision' fertilize parameter.

This allows you to let the amount of fertilizer you apply depend on
the nitrogen content of the soil in a specified zone.  Use it like
this: 

  (fertilize (n50)
             (precision (target 150.0) ; kg N/ha
                        (minimum 50.0) ; kg N/ha
                        (from 0.0) (to -100.0)))

Daisy will measure the amount of fertlizer in the specified zone (from
the soil surface to 100 cm depth), and supply enough extra nitrogen to
reach the specified target (150 kg N/ha).  If less than minimum (50 kg
N/ha) nitrogen is needed in order to reach the target, no fertilizer
will be applied.

You can use the 'precision' parameter with organic fertilizer as well,
in that case you must also specify the 'first_year_utilization'
fertilizer parameter. 

* 1.43

** New 'set_subsoil_irrigation' action.

For irrigating directly in the soil.  Unlike the other irrigation
actions, this one specifies a permanent flux.  This means the
irrigation will continue until you explicitly turn it off by
specifying a flux of zero.

Use it like this

    (set_subsoil_irrigation 1.5 (from -5.0) (to -10.0)
                            (solute (NH4 0.1) (NO3 0.7)))

This will specify an irrigation flux of 1.5 mm/h to be incorporated in
the soil between 5 and 10 cm depth.  The water will contain 0.1 mg
NH4-N/liter and 0.7 mg NO3-N/liter

If you call 'set_subsoil_irrigation' multiple times, the last call
will have effect.

Use

    (stop_subsoil_irrigation)

to turn off subsoil irrigation.

** Changed irrigation solute units from g/mm to mg N/liter.

* 1.42

** 'irrigate_top' action changed to 'irrigate_overhead'.

The old command is still available for compatibility reasons, but
marked obsolete and will be removed in at future version.

Also, the 'irrigation_top' log variables in the 'default' bioclimate
model hgave been renamed to 'irrigation_overhead'.  The 'log.dai' file
has been updated accordingly.

* 1.41

** New 'NEWS' file.

Read this to keep track of Daisy.

** Changed default log file ending from '.tab' to '.dlf'.

Those log files that contains a header now end with '.dlf' instead of
'.tab'.  This affects the build-in 'harvest' log model, as well as all
the 'table' parameterizations found in 'log.dai'.  The 'PMSW'
parameterization uses the old 'table1' log model, doesn't have a
header, and still output to 'pmsw.tab'.

** SoilNO3 and SoilNH4 now have default values.

You no longer have to specify the initial nitrate and ammonium content
when using the 'default' column model.

By default, the nitrate concentration in the water will be half of the
WHO recommended maximum for drinking water, i.e. 50 mg NO3 per liter.
The total ammonium content will a tenth of that.

Suggested by: SH.

* 1.40 

** New 'repeat' action.

The 'repeat' action take a single unnamed argument, which itself
should be an action.  That action will be repeated from the start, as
soon as it is finished.  If you for example have defined a 'sbarley'
manegement strategy, and you want to keep growing spring barley for
ever and ever, you can specify the manangement like this

        (manager (repeat sbarley))

More likely, you will use a 'while' to limit the repetition.

        (manager activity
           (while (wait (at 1998 7 7 1))
              (repeat sbarley))
           stop)

This will keep growing spring barley until 7/7 1998.

** New 't' action.

The 't' action will do nothing, just like 'nil', but will never be
"done".  This means that

        (while t (activity sbarley stop))

will do the same as

        (activity sbarley stop)

** New 'am/cleanup' AOM pool.

Small AOM pools will be merged into the 'am/cleanup' pool instead of
being mineralized instantly.  This way we avoid mysterios "spikes" in
the mineralization log data when small AOM pools are discarded.  

** Fixed log after 'default' crop get ripe.

In the 'default' crop model, production and respiration halts after
the crop gets ripe.  However, the last active production and
repsiration values would continue to be logged.  Now, these are
zeroed.

Bug found by: SH.

** Surface irrigation didn't work.  Now it does.

Surface irrigation was ignored.  Now it is applied correctly.

Bug found by: Mette Lęgdsmand <Mette.Laegdsmand@agrsci.dk>

** Fixed numeric error in the 'default' bioclimate model.

The symptom was a rare assertion failure.

Bug found by: SH.

* 1.38

** New 'inorganic' column model.

Works like the 'default' model woth regard to water and soil
chemicals, but does not keep track of organic mater or nitrogen in the
soil or surface.  The crops will always have potential N uptake.

** 'Surface Water Balance' now logs production stress.

* 1.37

** New 'vegetation.dai' file, for permanent vegetation parameterizations.

** Fix 'PMSW' log, broken by changes in 1.36.

** Use 'S_external' to log incorporated stuff.

The 'N Balance' and 'Soil Chemicals' logs are updated to do this.

* 1.36

** Log at end of the day.

This means that the data in the daily log files now corresponds to the
day listed.  The used to correspond to the previous day.  Fixed
'daily', 'weekly', 'monthly' and 'yearly' conditions to always match
the last hour in the period.

* 1.35

** A zero stress factor now means no stress.

Previously it went maximum stress.  Specifically, the meaning of
'water_stress' and 'nitrogen_stress' has been inverted.  Zero is no
stress, one is full stress.

** SVAT model integration changes.

The 'PMSW' svat model give feedback
to the rest of the system, by calculating a 'production_stress' which
will be used instead of 'water_stress'.

* 1.33

** SVAT model integration changes.

The 'PT' parameter of the 'default' bioclimate model is gone, instead
there now is a 'svat' parameter.

** The 'crop' action fixed so it no longer will harvest before sowing.

The symptom was that a crop was skipped when started at the wrong
time. 

Bug found by: HS.

* 1.32

** Optional parameters of the 'crop' action wasn't optional.

Now they are.

Bug found by: HS.

* 1.30

** Fix compilation with Borland C++ 5.01.

* 1.29

** Default value for 'dispersivity' (6.0 cm).

** Default value for 'rs_min' is now 30.0 s/m.

* 1.28

** New 'permanent' vegetation model.

It is intended for column on a multi-column simulation where we have
very little knowledge, basically we just need to know the LAI at
different times of the year.

* 1.27

** Column specific weather.

It is now possible to have separate weather for each column.  You can
still have a global weather, it will be used as default for columns
where no weather have been specified.

* 1.26 

** Log files now print a header.

Specifically, the 'harvest' and 'table' log models will now -- by
default -- create a header containing information about the data.

** New 'simple' crop model.

Intended for crops where we have no knowledge of reactions to water or
nitrogen stress.

* 1.24

** New Bentazon, IPU, and MCPP parameterization in 'chemistry.dai'.

** The 'crop' action now supports spraying.

** Added support for ridging.

* 1.23

** Log and harvest dead leaves.

* 1.22

** Added domain and range for PLF's in reference manual.

* 1.21

** Renamed CSMP to PLF.

* 1.19

** New 'fixed' and 'deep' groundwater models.

* 1.18

** New 'set_heat_source' action.

Use this to make a football field green a larger part of the year.

* 1.17

** New 'set_porosity' action.

Use this to change the porosity of the soil, to simulate the effect of
applying heavy machinery on the field.  Works only with the
'M_vG_compact' hydraulic model. 

** New 'M_vG_compact' hydraulic model.

This one has variable porosity.

** New 'set_detention_capacity' action.

This one is intended for simulating changed to the soil surface
ruggedness. 

* 1.14 

** New 'AOM' log model.

To keep track of a specific pool of added organic matter.

* 1.10 

** Use 'volatilization' instead of 'NH4_evaporation' in AM.

* 1.08

** Made heat and moisture influence user setable.

Affects mineralization, nitrification, and denitrification.

* 1.07 

** Made soil chemicals mix with surface chemicals.

* 1.06

** Added bioincorporation of organic matter from surface to soil.

* 1.05 

** New 'crop' action.

This action specifies a strategy for managing a specific crop or
multicrop.  To create a crop rotation, you can combine these 'crop'
action sequenctially using the 'activity' action.

* 1.04

** New 'lysimeter' groundwater model.

* 1.00

** New 'max', 'min' and 'average' selections for the 'table' log model.

** New 'assert', 'message', 'warning', and 'error' actions.

* 0.99

** Added surface runoff.

Controlable with the new 'DetentionCapacity' and 'ReservoirConstant'
surface parameters. 

** Added mixing of inorganic nitrogen between surface and soil.

* 0.96

** New 'fodderbeet' crop parameterization.

** New 'default' weather model.