Skip to content

Latest commit

 

History

History
680 lines (464 loc) · 38.5 KB

CHANGELOG.md

File metadata and controls

680 lines (464 loc) · 38.5 KB

Changelog

USGS Modflow One-Water Hydrologic-Flow Model (MF-OWHM)

New Feature Changelog

Boyce, S.E., 2023, MODFLOW One-Water Hydrologic Flow Model (MF-OWHM) Conjunctive Use and Integrated Hydrologic Flow Modeling Software, version X.Y.Z: U.S. Geological Survey Software Release, https://doi.org/10.5066/P9P8I8GS

Boyce, S.E., Hanson, R.T., Ferguson, I., Schmid, W., Henson, W., Reimann, T., Mehl, S.M., and Earll, M.M., 2020, One-Water Hydrologic Flow Model: A MODFLOW based conjunctive-use simulation software: U.S. Geological Survey Techniques and Methods 6–A60, 435 p., https://doi.org/10.3133/tm6A60

 

[[TOC]]

 

2.3.1-b3

TBA

git commit log: git log --reverse 455800a6bf6138e1b00c5ee17b6eb2108614e2e3^..HEAD

HEADER1

abc

Fixed

  • FMP Farm Process

    • SFR is not required when using the SURFACE_WATER block options: SEMI_ROUTED_DELIVERY, SEMI_ROUTED_RETURN, or ROUTED_RETURN_ANY_REACH, ROUTED_RETURN_ANY_NON_DIVERSION_REACH. Instead, if SFR is not part of the simulation, they are disabled a warning is raised.
    • BARE_PRECIPITATION_CONSUMPTION_FRACTION removed from FMP_Template.
      • The Farm Process input template, doc/FMP_Template/FMP_Template.fmp, incorrectly had a legacy keyword. This option was superseded by the SOIL block EFFECTIVE_PRECIPITATION_TABLE and the CLIMATE block PRECIPITATION_POTENTIAL_CONSUMPTION, which are applied to both bare and non-bare land uses.
    • Deep percolation conversion to runoff if all model cells are IBOUND=0 beneath it.
      • If all the cells beneath the land surface are IBOUND=0, then deep percolation is shifted to runoff. This was correctly implemented for the solver, but not for the output files. This should improve some of the reported mass errors in a simulation, but will not change the actual head solution.
    • Keywords that support optional List-Array Input (LAI) failed to identify the LAI keyword if it was CONSTANT or INTERNAL.
      • Several FMP keywords may only be the keyword itself to apply to all records, or optionally specify a LAI[S,T,L] input to specify the input on a record by record basis.
      • This effected:
        NO_RETURN_FLOW
        ZERO_CONSUMPTIVE_USE_BECOMES_BARE_SOIL
        EVAPORATION_IRRIGATION_FRACTION_SUM_ONE_CORRECTION
        ALLOCATION_BENEFICIARY_SPLIT
      • For example, NO_RETURN_FLOW [ LAI[S,T,L] ] indicates that all runoff should be converted to deep percolation for all Farms/WBSs. The following is an example, where the last two bullets would not pick up the LAI keyword but now does (assumes NWBS=3):
        • NO_RETURN_FLOW # Applies to all WBS
        • NO_RETURN_FLOW STATIC LIST OPEN/CLOSE no_return_flow.txt # input located in no_return_flow.txt
        • NO_RETURN_FLOW CONSTANT 1 # Applies to all WBS - Now works
        • NO_RETURN_FLOW INTERNAL # Specify for each WBS - Now works
          1 1 # WBS 1 has it enabled
          1 0 # WBS 2 has it disabled
          1 1 # WBS 3 has it enabled
    • Cleaned up information written to the list file that describe the Groundwater Allotment that is assigned.

  • BAS Basic Package

    • PRINT_WATER_DEPTH output file wrote an empty line to PRINT_WATER_TABLE output file.
      • Fixed PRINT_WATER_DEPTH from using the unit number associated with the BAS option PRINT_WATER_TABLE when writing an empty line to separate records. If PRINT_WATER_TABLE, then a random file called fort.xxx with xxx being a random number, would be written with nothing but blank spaces.

  • NWT Newton Solver (MF-OWHM specific version)

    • Allow keyword CONTINUE after the SPECIFY keyword.
      • The NWT solver instruction manual defines that a set of numbers are read after the SPECIFY keyword option. However, the MODFLOW-NWT source code allows for the CONTINUE keyword option to appear after SPECIFY and before the numerical parameters. This is also the default behavior for FloPy when making a NWT solver file. This feature was added back to maintain compatibility with MODFLOW-NWT and FloPy.
      • It is recommended to use the BAS option NO_FAILED_CONVERGENCE_STOP, which has the same effect and is not solver dependent.

  • SWO Surface Water Operations

    • Improved warning message descriptions.
    • RESERVOIR_SPLIT_FRACTIONS keyword added.
      • In an early beta release of SWO, the split reservoir setting required the keyword RESERVOIR_SPLIT_FRACTIONS, however for the final release of SWO this keyword was changed to RESERVOIR_PRIMARY_REREGULATION_SPLIT. For backward compatibility, both keywords are now checked for and serve the same purpose.

  • Slang MF-OWHM Language Interpreter (only used by SWO)

    • SFR.seg.INFLOW and SFR.seg.OUTFLOW are now correctly set for iteration-based slang scripts.
      • Slang can be invoked at different times of the simulation, model start, stress period start, time step start, iteration start, and at the end of those same times. For scripts invoked at the start of each iteration, the variables that are set to the inflow or outflow of a specific SFR segment reused the solution from the second iteration. It now updates with the solution of the previous iteration.
    • Add check for valid REQ.DELIVERY.FARM variables.
      • Slang property variables are set to values within MF-OWHM based on the current model state. The variables REQ.DELIVERY.FARM.# and REQ.DELIVERY_VOL.FARM.#, where # is replaced by a valid FARM/WBS number, are set to the water delivered to the WBS. However, if the user specified a number <1 or >NWBS resulted in random behavior, so a check was added to prevent this.
    • Fixed incorrect handling of duplicate variable names.
      • If a variable name, in the variable definition input is specified twice, then only one value should be stored. However, if a user happen to define the same variable name in the PROPERTY and RETURN variable blocks, then slang would correctly drop the duplicate from the PROPERTY block, but incorrectly drop the last variable defined in the block. However if the user entered a bad number, then either the variable was set to a random value or resulted in a runtime index error.

  • CFP Conduit Flow Process (MF-OWHM specific version)

    • Fixed an input read error for the advanced CFP input, which reads in a set of boundary condition flags to modify the input structure. If the advanced input is skipped, then an X must be used as a placeholder. However, this raised an error that the option was not found.
    • Index error for Time-Dependent Boundary Conditions (TD)
      • The search algorithm for finding the time interval no longer raises an index error if the time falls within the first interval.
      • The algorithm was also refactored to improve the execution speed.
    • Remove potential of taking sqrt of a negative number.

  • validate_example_results.f90

    • Fixed missing character initialization.

    • Includes owhm-v2 simple examples as part of validation.

  • examples/bash_example_run/1_RunValidation.sh:

    • The example driver bash script runs all the example problems and then runs a Fortran tool to check if they match a known hash value. If one or more examples do not match, then the Fortran tool returns a non-zero exit status code. However, on some versions of bash this would instead exit the entire script preventing the clean up and closing comments from running. Instead a set of if and conditional || are used to check if the examples pass and then the script completes normally.

  • options.print_convergence.txt fixed an incorrect header description.

Refactoring

  • FMP

    • EFFICIENCY keyword changed to IRRIGATION_EFFICIENCY
    • EFFICIENCY_IMPROVEMENT keyword changed to IRRIGATION_EFFICIENCY_IMPROVEMENT
      • The original keywords are still supported to maintain backward compatibility, but the new versions are now used in the FMP_Template and LIST output.

  • LPF and UPW improved wettable description.
    The LPF/UPW packages output the layer wettable description to the list file as NEVER-DRY or WETTABLE.
    To be more descriptive the variable now outputs:

    • NEVER-DRY
      • Indicates layer is confine
    • STAY-DRY
      • Indicates layer is convertible without wetting
    • WETTABLE
      • Indicates layer is convertible with wetting

  • fmp_template.fmp

    • Several of the FMP keywords did not indicate the units of the input. This has been corrected to include the units for all keywords.

  • CFP changed automatic arrays to allocatable to remove potential of a stack overflow for large models.

  • CFP changed exponents from real to integers to improve speed.

    • For example, x**3.0 was changed to x**3 to tell the compiler to expand the variable rather than using a floating point pow library. That is, x**3 results in the compiler doing x*x*x in the assembly code.

  • LineFeed comment cleaning and changed Fortran equality letters to symbols (.EQ. was changed to ==; .GT. was changed to >; ...)

  • Slang added to TYPE(VARIABLE_NAME_MEANING) the routine SUBROUTINE EQUALITY_VARIABLE_NAME_MEANING, which is assigned as the GENERIC :: OPERATOR(==) to allow variable equality checks.

  • Bash scripts now start with #!/bin/env bash instead of #!/bin/bash

 

2.3.0

2024-01-10

git commit log: git log --reverse 9d9f5b50c77a03b538e4ec818f5a67e7bcf3e5ea^..455800a6bf6138e1b00c5ee17b6eb2108614e2e3

HYDFMT v1.2

The update includes the MF-OWHM read utilities, new options, and fixed minor bugs.
See CHANGELOG_Features.md for a listing of new Hydmod and HYDFMT features.

ZoneBudget v3.3

Update to ZoneBudget from 3.2 to 3.3
The update includes the MF-OWHM read utilities, new options, and fixed minor bugs.
See CHANGELOG_Features.md for a listing of new ZoneBudget features.

Merge ⯬ MF-NWT v1.3

Merge ⯬ BiF v1.2.0

CFP MODE 2 and 3 is Disabled

  • CFP input specifies MODE as being 1, 2, or 3.
    • MODE 1 solves the conduit pipes (available)
    • MODE 2 solves conduit layers (disabled)
    • MODE 3 combines 1 and 2 (disabled)

Mode 2, and consequently 3, is failing to converge for the example problems nor matching the example results. In an abundance of caution this feature has been disabled until the reason for the discrepancy is found. If mode 2 or 3 is selected then an error message is raised.

Only MODE 1 is allowed until this issue is fixed.

Fixed

  • RCH reverted back to using upper most IBOUND /= 0 cell to match results from MF-NWT.

    • This removes the RCH fix in 2.0.1 that changed the check to include for convertible layers a check for if the head is above the cell bottom.
    • Added NRCHOP = 4 and NRCHOP = 5 as options to mimic the old behavior.
    • For details about this change, please see CHANGELOG_Features.md

  • FMP Bug Fixes

    • Can use FMP without SFR.
      • Previously, if FMP is used without SFR, then an index error occurred when FMP attempted to check for runoff and delivery locations. To prevent this index error from occurring the user had to specify the WATERSOURCE keyword to indicate that SFR is not available.
      • Now if SFR is not in use, then FMP automatically disables SFR diversions and runoff is flagged for flowing out of the model (rather than to a specific SFR location).
    • MNW2-FMP Supply well link no longer crashes if row and column do not match (SUPPLY_WELL Block). Instead, FMP now copies the row and column specified in MNW2. The row and column input is still required in FMP as a placeholder, but not used.
    • SURFACE_WATER block PRINT SFR_RETURN keyword with binary output (BINARY keyword) wrote a random integer if the WBS did not have assigned any SFR return flow locations.
      • The fix changed this to write the segment and reach as zero to indicate no SFR return flow point was specified.
      • Note this is how the text version of the option already works.
    • FMP failed to identify other blocks when LAND_USE was not specified in the input.
      • While this block should always be specified, the input indicates it is optional.
        The code now reflects this flexibility.
    • Fixed errors when NOT including the LAND_USE block in the input and:
      • CLIMATE block specified REFERENCE_ET to result in bare soil evaporative fallback calculations (fixed allocation error).
      • Raise an error when NCROP > 0
    • Fixed errors with FMP LAND_USE block and Bare Soil:
      • Improved logic for bare soil evaporation when NCROP = 0

  • Surface Water Operations (SWO)

    • Fixed an index error that occurs when the Slang input does not declare any required flow variables.

    • Budget routine missing initialization for 3 diversion accounting variables.

  • BAS Options Fixes

    • PRINT_HEAD, PRINT_WATER_TABLE, and PRINT_WATER_DEPTH no longer sorts the stress period and time step (SPTS).

      • Previously, in MODULE BAS_OPTIONS_AND_STARTDATE, if multiple PRINT_HEAD, PRINT_WATER_TABLE, or PRINT_WATER_DEPTH keywords were included to indicate output for different SPTS, then they would be sorted by stress period and time step. However, this only slowed the runtime rather than improved it.

    • PRINT_WATER_TABLE and PRINT_WATER_DEPTH file header write error.

      • In bas.f the subroutine GWF2BAS7OT checks if PRINT_WATER_TABLE and/or PRINT_WATER_DEPTH options are in use and if the corresponding arrays should be written to a file for the stress period and time step that just finished. While writing the header to the output, the code would use the file unit number associated with PRINT_HEAD option, which would either put an extra header in those files or raise a random access violation error.

    • Convergence output files changed date format to ISO Standard. This effects the following options:

      • PRINT_ITERATION_INFO
      • PRINT_CONVERGENCE
      • PRINT_FLOW_RESIDUAL
      • PRINT_RELATIVE_VOLUME_ERROR

  • SWR Bug Fixes

    • Reach ending layer assignment is set to 1 if the elevation is above the top of layer 1 and set to NLAY if it is below the bottom of NLAY. Previously, this situation resulted in the layer assignment being undefined.
    • The code block that updates the number of QAQ connections (NQAQCONN) was specified too early in the code. The CQAQCONN loop was moved to the correct location.

  • NWT Bug Fixes

    • Isolated model cells (ones that are surrounded by IBOUND=0 cells) were previously set to HDRY and the IBOUND changed to zero. To be consistent with other flow packages, the head value is instead changed to HNOFLO.

    • SUBROUTINE XMD7DA added declaration for IGRID argument rather than using implicit typing.

    • xmd_lib.f reorder declaring variables.

      • Several routines in xmd_lib.f declare the subroutine arguments after local variables. Most compilers work around this, but it can cause strange effects for arguments that represent local array dimensions.

      • For example:

        subroutine array_routine(dim)
     integer, dimension(dim):: work
     integer:: dim
end subroutine

        Has the issue of dim being used as an array dimension before being declared. Most compilers, will catch this and automatically use dim appropriately, but it is best to restructure the routine as:

        subroutine array_routine(dim)
     integer:: dim
     integer, dimension(dim):: work
end subroutine

  • HOB now requires drawdown observations to be in chronological order or an error is raised.

    • Either the user will need to fix the order or change the observation order or change to head observations.

  • MULT Bug Fixes

    • ExpressionParser inline IF conditional index error.

      • If the ExpressionParser solved an expression within an inline IF that did not result in all the values being True or all False (that is changing the entire array), then an index error was raised.

      • Inline IF is defined as follows:
        IF[ COND, trueANS, falseANS ]
         
        where:
        COND is a conditional expression (must contain <, <=, >, >=)
        trueANS is the result returned where COND is True
        falseANS is the result returned where COND is False.
                            It is optional to include falseANS,
                            if not present, then where COND is False,
                                then the result returned is NaN.  

      • For example, given
        x = 10.
        y = 5.
        would have:
        IF[ x + 3 > y, 2*x, 10*y ]
        return the trueANS, which is2*x or 20

        Another example is
        IF[ A < -1e-6 & A > 1e-6, 0, B / A ]
        would return an array with values of zero where A is near zero,
        otherwise sets the location to B divided by A

Refactoring

  • LPF and UPW expanded comment support by using MF-OWHM read utilities.

    • Provides better error and warning messages for bad input.
    • Input allows comments and empty lines.
      • This mainly affected when properties were defined by parameters and
        then the input would use a Fortran list-directed read to load the print factor (IPRN).
        In particular, the following code:
        READ(IN,*) LAYFLG(1,K)
        was changed to:
        CALL READ_TO_DATA(LINE, IN, IOUT)
        LLOC = 1
        CALL GET_INTEGER(LINE, LLOC, ISTART, ISTOP, IOUT, IN, LAYFLG(1,K), ...
    • DIS improved end of file error message when not enough stress period information is specified.

  • FMP CLIMATE Block refactored code to allow specifying DIRECT_RECHARGE multiple times and the sum of the recharge arrays are applied to deep percolation.

  • main.f90 white space cleanup and reordering of HYD subroutine calls.

    • Fixed indentation to make code line up.
    • Modified an if statement to improve speed when using HYD and calling GWF2HYD7STR7RP and GWF2HYD7SFR7RP routines.

  • xmd.f indentation cleanup.

  • tabfile_module.f reading the actual tabfile's filenames now accept the keywords OPEN/CLOSE, DATAFILE, and DATAUNIT.

    • Previously, the tabfile filename was specified without a keyword as just FNAME or EXTERNAL UNIT.
      If the user used a keyword, such as OPEN/CLOSE ./myTabFile.txt,
      then the program would stop saying Failed to open "OPEN/CLOSE" causing confusing to the user.

  • util.f subroutines no longer raises a warning if the optional inputs CNSTNT or IPRN are not provided.

    • This effects the following subroutines:

      • U1DREL
      • U2DINT
      • U2DREL
      • U2DDBL

    • Instead now, if either is not present, the following is written to the LIST file:
      Reading for XYZ, autoset the scale factor CNSTNT to 1.0, and the print flag IPRN to -1

  • WEL and MNW2 improved missing ITMP warning.

    • For the MNW2 and WEL packages, if the end of file is reached or the input fails to read ITMP, then ITMP is assumed to be zero for the rest of the simulation.
    • When this occurs a formal warning was made.
      This was changed to being a minor message in the LISTing file with the words:
      Don't Panic
      in the warning message.
    • It also indicates to the user that if they are using LINEFEED or FMP-Link the rates may be initialized outside of the package.

  • FMP variable DRTFLOW is always allocated when using FMP.

    • The DRT-FMP link requires the intermediate variable DRTFLOW. This variable was only allocated when DRT and FMP are both in use. To simplify the connection, FMP now always allocates DRTFLOW when FMP is in use.

  • FMP improved "runoff leave model" warning message.

    • If a WBS/Farm contains runoff, but has no where to go, a soft warning is raised to let the user know which WBS has runoff and that the runoff is flowing out of the model because it has no where within the model to go.

      The warning has been expanded to include suggestions on how to fix the problem and what potential causes of the warning are.

      The warning also indicates that water is leaving the model domain automatically since the user did not specify a runoff location.

  • FMP source file standardization of indentation for the following files:

    • src/fmp/allotment_data.f90
    • src/fmp/climate_data.f90
    • src/fmp/crop_data.f90
    • src/fmp/options_data.f90
    • src/fmp/output_data.f90
    • src/fmp/salinity_data.f90
    • src/fmp/soil_data.f90
    • src/fmp/surface_water_data.f90
    • src/fmp/surface_water_operations_data.f90

  • src/fmp/wbs_data.f90 added SETUP_BASIC_VAR_WBS_DATA subroutine in WBS_DATA_FMP_MODULE

    • In module WBS_DATA_FMP_MODULE the basic variable initialization and allocations done by SUBROUTINE INITIALIZE_WBS_DATA and SUBROUTINE SETUP_NO_WBS_DATA are moved to SUBROUTINE SETUP_BASIC_VAR_WBS_DATA.

    • This minimizes the redundant code between the two routines.

  • src/fmp/crop_data.f90 moved crop initial array allocations to separate subroutine.

    • The FMP LAND_USE block, MODULE CROP_DATA_FMP_MODULE, contained conditional array allocations that occurred in the SETUP_NEXT_STRESS_PERIOD subroutine (RP routine). These have been moved to SUBROUTINE SETUP_DEPENDENT_PARTS.
    • By doing this, the allocations can occur at the FMP allocate and read routine (SUBOURITNE FMP_AR) after all the FMP block inputs have been read.

  • hydfmt.f minor character variable cleanup to reduce final executable binary size.

    • To reduce the size of the binary, variables that were set to the same static string only set the first variable and then the remaining were set to the first.
    • For example:
            a = '---------'
            b = '---------'
            c = '---------'
      is changed to:
            a = '---------'
            b = a
            c = a

 

2.2.0

2022-01-20

git commit log: git log --reverse 4bfb023b3a0f18d8a53a35146f85a93528d6ddd0^..9d9f5b50c77a03b538e4ec818f5a67e7bcf3e5ea

ZoneBudget v3.2

Update to ZoneBudget from 3.01 to 3.2
The update includes the MF-OWHM read utilities, new options, and fixed minor bugs.
See CHANGELOG_Features.md for a listing of new ZoneBudget features.

Merge ⯬ BiF v1.0.2

Fixed

  • NWT Bug Fixes

    • NWT considers the MODFLOW Outer Iteration (1 to mxiter) to be each time it solves for the Jacobian, and does not include iterations solved with Backtracking (residual control). This causes a disconnect with the other MODFLOW packages that assume an outer iteration occurs every time the aquifer flow equations are formulated (package's FM routines). To fix this issue, the number of MODFLOW Outer Iterations is always accounted for as the number f times the FM routines are called.

      As a result of this, previous models may need to increase their maximum number of NWT iterations (MAXITEROUT) to allow for convergence. Previously, the number of outer iterations was under-reported such that a model that said it converged in XX iterations, really converged in YY iterations, where YY >> XX.

      Please see the CHANGELOG_Features.md section 2.1.1 Section NWT Improvements for detailed information about this.

    • NWT thin cell check is now disabled by default.

      • To enable the thin cell check, add the option THIN_CELL_CHECK.
        All "thin" cells removed from the simulation are written to the LIST file.
  • SFR Modifications
    • The flow-depth-width lookup table, ICALC = 4, use log10 interpolation,
      except when the flow rate is less than the first point, then it uses linear interpolation.
      If the first lookup point is <1.0E-30, then SFR uses linear interpolation for the first two points.
      • This prevents a floating point overflows from doing log10(0.0).

  • FMP Bug Fixes

    • EVAPORATION_IRRIGATION_FRACTION BY_IRRIGATE LIST keyword changed to read a list of NIRRIGATE fractions instead of the incorrect NCROP fractions.
      • The keyword EVAPORATION_IRRIGATION_FRACTION BY_CROP LIST is how to read NCROP fractions.
    • Fixed issue when defining NON_ROUTED_DELIVERY VOLUME STATIC and NON_ROUTED_DELIVERY STATIC that would continue to reduce for each stress period the available NRD water.
      • FMP would read the NRD volumes once, divide the volume by the stress period time to get a rate, then for each subsequent stress period divide the previous rate by the next stress period's time. Now, the rate is determined using the original volume defined.
    • Added checks for when NCROP = 0 to allowing running FMP without any land use/crops defined.

  • MNW2 Bug Fixes

    • Previously, the well head (hwel) was set to cell head for wells that went dry
      (that is, pumping that exceeded the well capacity).
      • Now hwel is set to the bottom of the well when it goes dry.
      • This affects the actual pumping for wells with partial penetration (PP) enabled in a convertible layer.
    • MNW2 now only does the RHS formulation for the last 10 solver iterations in a time step that does not converge.
      • That is, if the solver has MXITER = 200 iterations,
        then MNW2 solves for odd iterations by modifying the RHS (WEL-like action)
        and even iterations by modifying the HCOF (GHB-like action),
        then only modifies RHS for solver iterations ≥190.
      • This improves the mass error for time steps that fail to converge,
        but have an option to continue to the simulation.
        For example, BAS option NO_FAILED_CONVERGENCE_STOP or NWT CONTINUE option.
    • MNW2 now recalculates the node flows if the time step converges on an even iteration.
      On odd iterations MNW2 specifies the node flow rate directly by modifying the RHS, but
      for even iterations it uses the previous iterations HNEW to modify HCOF
      and make the node flow a function of the next iterations HNEW.
      • The difference between the last two iterations of HNEW can significantly change the reported node flows.
    • MNW2 and NWT fixed issue when writing to the cell-by-cell for nodes that have gone dry because NWT/UPW does not change the IBOUND array.
  • SWT now zeros out the BUFfer array before using it fixing an issue during Steady State stress periods.
    • Previously, SWT would write what the previous package had stored in BUF causing it to report flows that did not exist.

  • SUB fixes

    • Inelastic compaction is assumed to be complete if the head falls below the cell bottom.
      • That is, if the critical head falls below the cell bottom, then it is set to -3.40E+38 to represent -inf.

    • Compaction (inelastic and elastic) and expansion (elastic) only occurs
      when the cell head is greater than the cell's bottom elevation.

      • Assume HOLD is the time step starting head, HNEW time step ending head, BOT is the cell bottom:
        if a time step has HOLD>BOT and HNEW<BOT, then compaction occurs but solves with HNEW=BOT.
        if a time step has HOLD<BOT and HNEW>BOT, then expansion occurs but solves with HOLD=BOT.
        if a time step has HOLD<BOT and HNEW<BOT, then nothing happens.

    • Changed the use of the PACK function to to use a mask array that is the same dimensions as BUFFER array. Previously, the mask was set to the scalar variable TRUE. This can have different meanings on different compilers.
      An example of the changed code is:
      RNB(LOC1:LOC2) = PACK(BUFFER, TRUE) was changed to:
      RNB(LOC1:LOC2) = PACK(BUFFER, PACK_MASK)
      where real, dimension(NCOL,NROW):: BUFFER
      logical:: TRUE = .true.
      logical, dimension(NCOL,NROW):: PACK_MASK = .true.

  • CHOB now works when using the UPW flow package and NWT solver.
  • U2DREL with BINARY input option now raises a warning instead of stopping with an error.
    • When the BINARY option is found, MF-OWHM attempts to identify the binary input structure.
      If it succeeds, a warning is raised about the portability of binary-unformatted files.
      If it fails, then the program stops with an error message.
  • expression_parser.f90 improved error message that is raised when a variable name is similar to a reserved keyword.
  • linefeed.f90 now raises an error if the end of a FeedFile is reached before the end of the simulation.
    • A proper FeedFile should have one line of input per stress period simulated.
      Previously, if the end of the file was reached, then the last lines input would be reused.
  • Removed from the visual studio solution (ide/visual_studio/OneWater_Project.sln) the key entry:GenAlternateCodePaths="codeForAVX"
    • AVX acceleration caused floating point truncation that resulted in model results that were not repeatable. That is, the same model input would produce different numbers after the tenth digit. Results are now identical when running the multiple times with the same input.
  • Multiple files changed floating literal numbers that did not include a decimal point to have one.
    • For example, X = 0D0 sets X to the integer 0, rather than double precision 0.0. The compiler than changes the zero from the integer to a double on the fly, but at the minor expense to runtime. To be syntactically correct, this is changed to X = 0.D0 to indicate a double precision number.

Refactoring

  • SFR merged SUBROUTINE GWF2SFR7LAKE into SUBROUTINE GWF2SFR7AD.

    • Routines were redundant and called at the start of the time step.
      Combining the two routines as one improves the readability of the code and has no affect on simulation results.

  • Added newly supported keywords to doc/Option_Block_Cheatsheets

    • BAS_Options_All.bas
    • BAS_Options_Recommended.bas
  • Added newly supported keywords to doc/Notepadpp_Syntax_Highlighting/userDefineLangs
    • BAS-DIS.xml
    • MF-Flow-Packages.xml
    • MF-Packages.xml
  • Changed DO CONCURRENT loops to regular DO loops for code with multiple branches (IF-ElSE).
    The current Fortran compilers seem to have issues with numerical accuracy
    for complicated statements in a DO CONCCURENT loops.
  • fmp_main_driver.f90 changed .NE. to /=
  • surface_water_data.f90 changed the NON_ROUTED_DELIVERY code from procedural to object oriented.
    The new object is TYPE NRD_VALUES and keeps track of consumed and available imported water based on assigned demand.
  • SUBROUTINE FMP3WELBD(KSTP,KPER,DELT,DATE,IGRID) for the FB_DETAILS and FB_COMPACT output files rearranged the code for clarity and execution speed.
    Changes occurred in IF(FMPOUT%FB_COMPACT%IS_OPEN .OR. FMPOUT%FB_DETAILS%IS_OPEN) THEN statement block.
  • BAS optionBUDGETDB , which writes for each time step the volumetric budget for each package writes the absolute value of the mass balance errors (always positive) for the PERCENT_ERROR column.

 

2.1.0 ⯬ MODFLOW Surface Water Operations (SWO)

2021-05-25

Initial release of MODFLOW Surface Water Operations (SWO) in MF-OWHM

  • Ferguson, I.M.., Llewellyn, D., Hanson, R.T., and Boyce S.E., 2016,
    User guide to the surface water operations process—An integrated approach to
    simulating large-scale surface water management in MODFLOW-based hydrologic models:
    Denver, Colo., Bureau of Reclamation Technical Memorandum no. 86-68210–2016-02, 96 p.

git commit log: git log --reverse d8ec82ae504a2aaec594ccd576f8674961f59404^..4bfb023b3a0f18d8a53a35146f85a93528d6ddd0

Fixed

  • UZF incorrectly set ZEROD9 = 1.0d0-9, now it is 1.0d-9 (that is, it was set to -8 and now it is 10-9).
    • This variable was used for near zero conditional checks.
  • GMRES Solver incorrectly set Stop_toldum = 1.0d0-9, now it is 1.0d-9.
    • Stop_tol_gmres is the tolerance for convergence of the iterative solver
  • MNW2 reported the well head (hwel) as the cell head for wells that went were dry (that is when pumping that exceeded the well capacity).
    • This affects the actual pumping if the well has partial penetration (PP) enabled and was in a convertible layer.
      Because it would apply the PP correction to what is a seepage face rather than a pumping well.
  • CHOB now works when using the UPW flow package and NWT solver.
  • U2DREL with "BINARY" input option now raises a warning instead of stopping with an error.
    • When the "BINARY" option is found, MF-OWHM attempts to identify the binary input structure.
      If it succeeds, a warning is raised about the portability of binary-unformatted files.
      If it fails, then the program stops with an error message.

Refactoring

  • Added slang/s_language.f90
  • Added slang/s_language_global_pull.f90
  • Added fmp/surface_water_operations_data.f90

 

2.0.3

2021-05-25

git commit log: git log --reverse 3adf1e3b8e634d83b8296fd673b3e3360a5cae06^..d8ec82ae504a2aaec594ccd576f8674961f59404

Fixed

  • Improvement on warning and error messages in MNW2, UPW, and NWT packages.

 

2.0.2

2021-05-15

git commit log: git log --reverse 12b331ce38c47a7e88f7da234c189ffa585d637a^..3adf1e3b8e634d83b8296fd673b3e3360a5cae06
or web view at: https://code.usgs.gov/modflow/mf-owhm/-/compare/2.0.1..2.0.2

Merge ⯬ BiF v1.0.1

Fixed

  • SFR issue for segments with ICALC ≥ 2 that would solve for stream depth with Newton–Raphson with an initial stream flow guess of zero instead of the stream reaches inflow.
    • For most cases this fix only affected simulation runtime and only altered the solution after the sixth significant digit (single precision tolerance).
  • FMP — LAND_USE (Crop) Block output files did not correctly write TOT_SURF_RUNOFF
    for crops that set SURFACEWATER_LOSS_FRACTION_PRECIPITATION to one.
    • The affected options were BYWBS, BYWBS_BYCROP, BYCROP, ALL, and ALL_VERBOSE
    • This error had no affect on the actual simulation results nor other output files.
  • MNW2 using the THIEM losstype resulted in a warnings being triggered for a near zero skin radius, which is an input option used by he SKIN losstype.
  • MNW2 using NWT resulted in the specific storage was not calculated correctly for use in the partial penetration correction.

 

2.0.1

2021-03-15

git commit log: git log 12b331ce38c47a7e88f7da234c189ffa585d637a

Merge ⯬ MF-NWT v1.2

Fixed

  • BCF was not recognized as one of the supported packages.
    While not recommended, BCF is supported.
  • SFR issue that causes it to use one layer deeper then the water table layer when all layers are defined as convertible.
  • MNW2 using QLIMIT with NWT resulted in the models that failed to converge do to a bad index reference for well head.
  • RCH and NWT packages with NRCHOP=3 did not pass recharge to the time step's upper most active layer. Previously, it only passed water to the upper most non-zero IBOUND cell rather than the upper most non-dry cell.
    • To mimic the original behavior of RCH with NWT set NRCHOP = -1, which applies recharge to the initial upper most non-zero IBOUND cell.
    • Removed in 2.3.0, see CHANGELOG_Features for more information.
  • UPW/NWT packages with convertible layers kept releasing water from storage after a model cell was dry.
  • HydMod issue with HD (head) interpolation used the same four points for all observation points, which resulted in an extrapolation. Fixed such that head observations are interpolated by creating a four point finite element from the four closest cells to interpolate for the requested X,Y head location. (this is similar to how Hob operates).

Refactoring

  • Source code organized and renamed for clarity and
    for some of the source code the package main module was moved to a separate file.
    • gwf2ghb7_OWHM.f is now ghb.f
    • gwf2mnw27.f is now mnw2.f and mnw2_module.f90
  • Minor spelling corrections in warning and error messages.
  • Added path_interface.f90, which provides subroutines for building Windows and Unix Paths and allows creating directories.
  • The modules in all_util.f90 were split into 48 separate files. The file name pertains to the util module it contains.

 

2.0.0

2020-04-07

Initial Release of MF-OWHMv2.

For a full listing of changes in this release see:

  • Boyce, S.E., Hanson, R.T., Ferguson, I., Schmid, W., Henson, W., Reimann, T., Mehl, S.M., and Earll, M.M., 2020, One-Water Hydrologic Flow Model: A MODFLOW based conjunctive-use simulation software: U.S. Geological Survey Techniques and Methods 6–A60, 435 p., https://doi.org/10.3133/tm6A60

Refactoring

Naming convention of source files:

  • Is the name of the package or the name of the package plus _module to indicate it contains the packages global variables.
  • Is a generic name for the collection of subroutines within the source file.
  • _interface indicates source code contains a generic INTERFACE call for a set of subroutines for a specific task.
  • _instruction indicates source code defines one or more Derived Data Types definitions (Fortran Objects) and their associated methods (subroutines and functions associated with the object).