PHABSIM - Preface

Terry Waddle
U.S. Geological Survey
2150 Centre Avenue, Bldg C
Fort Collins, CO 80526

Citation: Waddle, T.J., ed., 2001, PHABSIM for Windows: User's Manual and Exercises: Fort Collins, CO, U.S. Geological Survey, 288 p.

Preface

Chapter 1

Chapter 2

Chapter 3

Chapter 4

Chapter 5

Appendices

Lab Exercises 1-6

Lab Exercises 7-12

Acknowledgments

This document evolved from IF310 course materials developed and revised numerous times since 1984. Several individuals made significant contributions to the Windows™ version of the software and this version of the manual in the following order. All are employees of the Fort Collins Science Center (FORT), U.S. Geological Survey (USGS), unless otherwise noted. Mr. John Bartholow wrote early versions of the laboratory exercises. Mr. Ken Bovee provided revision and insight to habitat suitability curve development and testing. Dr. Robert Milhous wrote the original PHABSIM programs. He also wrote the PHABSIM Version 2 (DOS) reference manual with assistance from Ms. Marlys Updike and Ms. Diane Schneider (Milhous et al., 1989). Substantial portions of this text and of the Version 2 class text were taken or adapted from that earlier work. Dr. Thomas Hardy, Utah State University, consolidated and expanded on earlier course materials to provide a working draft of the PHABSIM Version 2 (DOS) class text under contract to the FORT. Substantial portions of the text in this document (in all chapters and in the laboratory exercises) are adapted from that draft. Dr. Sam Williamson contributed numerous insights into the techniques and strategy of habitat modeling and wrote the descriptive text for those ideas. Mr. Jeff Sandelin designed and programmed the Windows™ interface and made numerous contributions to the text while employed under a support services contract. Ms. Julianne Brown, also a support services contractor, performed extensive testing of the programs to ensure results were correct and congruent with the previous version of the software. Mr. Jim Henriksen provided numerous insightful review comments and text revisions while using the PHABSIM for Windows programs. Dr. Terry Waddle supervised the design and testing of this software version. He also contributed substantial new text material while editing this version of the manual and laboratory exercises to reflect the overarching IFIM concept and the PHABSIM for Windows structure.

TOP OF PAGE

Preface

A Brief History of PHABSIM

In the 1970's a major change in terminology regarding maintaining streamflow to protect aquatic organisms in streams was introduced. To some, the change was considered merely semantic. But to those whose pioneering work led to the change, it was both substantive and significant. The change was from common usage of the term "low flow" or "minimum flow" to the common usage of the term "instream flow." During the early 1970's, the Water Resources Research Catalog served as an index of most of the ongoing research related to water resources, with each project identified by several key words. As late as 1975, the Catalog contained no research projects under the key words "instream flow," while at the same time, there were many studies using the key words "low flow."

The typical description of the concept of "low flow" in those days was something like this:

"Water is taken out of the stream for a variety of uses, such as irrigated agriculture, municipal and industrial. Low flow means that amount of water that must be left in the stream for the fish. With anything less than the low flow, the fish will die."

In a thoughtful piece published in the April 1991 edition of Rivers, Harvey Doerksen recalled how Don Tenant, then an aquatic biologist for the State of Montana, described the shortcomings of the "low flow" concept in human terms:

After many years, I still have a vivid recollection of a photographic slide that [Don Tennant] showed as part of his presentation at a conference. It was a picture of a family of perhaps half a dozen people crowded into a small bathroom. ... The point that Don wished to make was simply this: in the short run, a population of fish in the stream can make out quite well with an extreme low flow event, just as a family can tolerate the closeness that comes with crowding into a small bathroom. Over extended periods, however, the low flow could not be tolerated any better by a population of fish than sustained proximity could be tolerated by a family in a crowded bathroom. And yet, that very concept of maintaining a sustained low residual flow was the prevailing approach to protecting fish habitat.

One of the serious problems with the "low flow" approach was that biologists distinguished only between two relative conditions with respect to fish habitats: the level below which disaster would occur, and everything else. However, other water users made incremental assessments of their need. An irrigation district, for example, could project water quantity needs for any increment of irrigated acreage, but the biologists at the time did not have the technology to do similar incremental assessments of the potential impacts of various irrigated acreage scenarios.

In the late 1970's the U.S. Fish and Wildlife Service (FWS), through what then was its Office of Biological Services, received funding from the Environmental Protection Agency (EPA) to establish the Cooperative Instream Flow Service Group, which incorporated the combined talents of people from several federal and state agencies. The Group's central charge was to develop methods for quantifying the biological effects of altered streamflows.

The result of a concerted effort on the part of the Instream Flow Group was the Instream Flow Incremental Methodology (IFIM) (Stalnaker et al., 1995; Bovee et al., 1998), of which the Physical Habitat Simulation System (PHABSIM) is a major component. This new technology was highly significant for at least two reasons. First, it allowed fishery biologists to negotiate acceptable flow levels with other instream and out-of-stream water users from among a variety of possible scenarios, in the same way that the other users had been doing for years.

Second, the change in the prevailing terminology from "low flow" to "instream flow use" meant that biologists no longer were trying to find that magical flow level below which a stream should not be dewatered. Instead, they were in a position to assert instream flow needs for fish habitats and other environmental values. Furthermore, they could do so in terms of the seasonal life cycle needs of the fish (or other aquatic organisms) over the annual hydrograph. This change was thus an instrument through which fish and associated environmental values were viewed as legitimate water users among many, instead of merely a residual, after the water users had been served.

This change was not easy, nor did it occur overnight. In developing PHABSIM, the Instream Flow Group drew on several developments in instream habitat assessment that were available at the time. Two developments were of particular importance, the Washington Method (Collings et al., 1972) and the univariate curve concept. The Washington Method provided the concept of mapping depth and velocity conditions over gravel bars and applying binary suitability functions for salmon spawning in streams in the Pacific Northwest. The area of a gravel bar suitable for spawning was evaluated at several measured discharges by calculating the area having a suitability value of 1 for both depth and velocity. Approximate suitable spawning area for unmeasured intermediate flows was estimated by interpolation.

Binary suitability functions (the observed condition is assigned a value of 0 or 1, unsuitable or suitable) produced a value of 0 for some stream areas that, although not optimally suitable, were observed being used by salmon. The use of a univariate suitability function, which ranked various depths or rates of flow velocity on a 0 to 1 scale and allowed a smooth function covering the entire range of conditions, was proposed by Waters (1976).

The Washington Method, even when modified to include the univariate curve approach, requires numerous empirical measurements at different discharges. This limits the number of discharges for which habitat can be evaluated and the number of study areas that can feasibly be evaluated with time, manpower, and budget constraints.

The Instream Flow Group combined standard one-dimensional hydraulic simulation techniques with the Washington Method and the univariate curve concept to produce PHABSIM. PHABSIM uses the hydraulic simulation models to predict depth and velocity at unmeasured flows using basic physical and engineering principles that were standard practice in the 1970's. The resulting software suite multiplied surface area for a section of stream by the univariate suitability curve values for depth, velocity, and channel condition to arrive at a habitat index called Weighted Usable Area.

This software suite was first implemented on Control Data Corporation mainframe computers accessed by terminals connected over telephone lines. The software consisted of numerous small piecemeal programs that were cost-effective to develop and use in the mainframe environment. Numerous small files were required for data input and numerous output files were produced. This placed a major burden on users of the software to manage large numbers of files with often similar (but slightly different) contents. It also produced a legacy of program names such as IFG-4.

During 1984 and 1985 the PHABSIM program suite was moved to microcomputers. A basic menu-driven interface for the PHABSIM program was developed in 1989. The PHABSIM Version 2 (DOS) (Milhous et al., 1989) programs were distributed until September 2000 as the standard version of PHABSIM software.

With the advent of the Windows graphical user interface, it became evident that the PHABSIM program suite would serve a wider audience if input data and model results could be displayed graphically during an application of the software. Development of the Windows interface for PHABSIM fulfills commitments made to the U.S. Fish and Wildlife Service in the mid-1990's. Two principle goals were pursued while developing PHABSIM for Windows: (1) to keep the functionality of all core programs and retain backward compatibility to PHABSIM Version 2, and (2) to simplify the bookkeeping and calibration processes. We sought to both improve clarity of the various options for each program and graphically display the results of applying each program. The ability to view plots of model results during the calibration and option selection processes greatly speeds up calibration and increases understanding of the effects of selecting various habitat simulation options.

To ensure the quality of PHABSIM for Windows, extensive testing and comparison with the DOS-based PHABSIM Version 2 was performed. Several data sets were run through both versions and all options of all programs were run and compared. In addition, both English and metric units were tested to ensure that entire analyses could be reliably conducted in metric units. Our emphasis was on retaining the full set of PHABSIM analytical functions and extensively testing the programs rather than on adding program enhancements.

Use of This Document

This document is a combined self-study textbook and reference manual. The material is presented in the general order of a PHABSIM study placed within the context of an IFIM application. The document may also be used as reading material for a lecture-based course. This manual provides documentation of the various PHABSIM programs so every option of each program is treated.

This text is not a guidebook for organization and implementation of a PHABSIM study. Use of PHABSIM should take place in the context of an IFIM application. See Bovee et al. (1998) for guidance in designing and performing a PHABSIM study as part of a larger IFIM application.

The document concludes with a set of 12 laboratory exercises. Users are strongly encouraged to work through the laboratory exercises prior to applying the software to a study. Working through the exercises will enhance familiarity with the programs and answer many questions that may arise during a PHABSIM analysis.

Changes Between PHABSIM Version 2 and PHABSIM for Windows

The first obvious change between PHABSIM Version 2 (DOS) and PHABSIM for Windows is use of the Windows graphical user interface and development of the necessary interface programs. The graphical user interface program now includes considerably more code than the analytical core programs. Additional benefits of developing an extensive interface program include the following: file management is largely done by program; users no longer need to keep track of file names and the risk of removing a file needed at a later step has been essentially eliminated and each study site is kept in a separate, user-specified, project directory, reducing the risk of confusing project files.

In developing this version of PHABSIM, certain decisions were made regarding the structure of PHABSIM data and analyses. All cell definitions now match the data gathering approach taught in the IF305 field techniques course. The cell definition distinctions HABTAE and HABTAV have disappeared and all HABTAV options are now included in HABTAE. Thus, there is no HABTAV program. For consistency, HABTAM cell definitions also match those used in HABTAE. The 100 cell per cross section limit has been eliminated so an unlimited number of cells may be defined for each cross section.

Another difference is the consolidation of AVDEPTH and AVPERM into one program. In doing so, the options and outputs of both programs have been retained.

With these changes have come some differences in computed results. The core analytical programs were translated into C++ from Fortran. Different compilers have different methods of creating machine language so there can be small differences in results of numerical calculations. This is largely due to a combination of order of the calculation, round-off error, method of optimization used within the compiler, etc. While in most situations the differences between the DOS PHABSIM Version 2 and PHABSIM for Windows are very small, there are situations where these differences can accumulate.

The following is an example of possible calculation discrepancies that may occur between the DOS PHABSIM and Windows PHABSIM. Both the Windows and old DOS versions were run for the same study site. The results for a specific cell in the study site at the same discharge and one selected life stage are shown in Table 1.

For illustration, the Weighted Usable Area (WUA) calculation had equal velocity Suitability Index (SI) and channel index SI values (1.0 and 0.91 respectively.) For the species/life stage in question, the depth suitability curve had a steep slope at this point (depth of 0.8 = SI of 0.06 and depth of 0.85 = SI of 0.15). The steep slope of the suitability curve gave a large difference in suitability for a small difference in calculated depth. Thus, noticeable differences in the final computed WUA value between DOS PHABSIM Version 2 and PHABSIM for Windows are likely in situations where the SI criteria are very sensitive to changes in the parameter(s) (depth in this case) for that cell.

Thus, for a relatively small difference of 0.014 in WSL, we show a large difference in WUA for the cell. The WSL model is performing as expected and is within reasonable calibration error for the data set. The differences in water surface elevation have been determined to be due to compiler differences as all calibration parameters were set to the same values in both versions and the internal calculations were traced to ensure they were being accurately performed.

It should be noted that over an entire cross section or over a study site, few cells show these dramatic differences and the resulting calculated WUA differences are usually less than 1 to 2%. Although no study site WUA difference greater than 10% was found during testing, larger differences could occur if a study encounters a situation where numerous cells have differences similar to that illustrated above.

Table 1. Example of propagating differences between PHABSIM for Windows and DOS PHABSIM.

Conventions Used in This Document

IOC References to Old PHABSIM

PHABSIM Version 2 (DOS) (Milhous et al., 1989) used lines of numerical switches to control program options. These were referred to as IOC (originally input/output control) lines. Users became familiar with selecting program options by remembering items like "IOC 5 set to 2 and IOC 8 set to 1". While these options are now described in the PHABSIM for Windows interface using text phrases, the old IOC numbers are also displayed with the text to aid experienced users in migrating to PHABSIM for Windows.

Menu Notation

The PHABSIM for Windows software is menu driven through a graphical user interface and the familiar Windows point and click sequence to operate program menus and options. An abbreviated notation convention is used throughout this document to instruct readers in how to make option selections without filling a large amount of space on the page with screen images for each action to be performed.

The following example illustrates how to assign the calibration data sets to each of the desired simulation flows when using the STGQ model. The first item is part of the main menu list that is always displayed at the top of the main PHABSIM window. Subsequent items are submenu items or choices that users will see on the screen as they work across the line. The "/" character is a delimiter to separate higher-level menu items from lower level items. Menu navigation notation is always displayed in bold type.

Example notation for using PHABSIM for Windows menus:

"To assign calibration sets to specific simulation discharges, go to /Models/WSL/STGQ Options/Assign Cal Sets and click the check boxes for the appropriate discharge/cal set combinations in the table shown".

To follow this example, the user would click on the main menu item Models followed by clicking in order the WSL item in a drop down menu, the STGQ Options tab, and the Assign Cal Sets button to arrive at the data entry table. This notation is further illustrated in Chapter 1.

TOP OF PAGE