|
<br />at the
<br />USGS
<br />i Place
<br />mge is
<br />Canal
<br />:rsions
<br />) pipe-
<br />
<br />itream
<br />at the
<br />:,from
<br />Joking
<br />r-view
<br />Ie, and
<br />ink, [Q
<br />'r mid-
<br />:Iuding
<br />ermine
<br />)w was
<br />Tound-
<br />
<br />which
<br />e scen-
<br />y color
<br />'ere not
<br />present
<br />Poudre
<br />jeo se-
<br />ld envi-
<br />at were
<br />ng and
<br />oks. but
<br />Itage of
<br />
<br />lPturing
<br />j to the
<br />: nearly
<br />eclining
<br />. Scenes
<br />,m 2643
<br />ptember
<br />'Is) later
<br />~ees had
<br />: vegeta-
<br />h earlier
<br />';
<br />
<br />..
<br />different ~
<br />)tunng a f',_
<br />.r, sunny
<br />: Canyon
<br />) always
<br />propriate
<br />
<br />, .
<br />
<br />BROWN ,....ND~.~ PA10sCAPE AESTHETICS OF RIPARIAf'l ENVIROf'lMENTS
<br />
<br />1789
<br />
<br />flow levels. Also, because all 22 points could not be photo-
<br />graphed in one day, typically only 15 or Ihe 22 points were
<br />photographed at anyone How rate.
<br />Video sequences were taken for 30 s on 314-inch film with
<br />a Sony DXC-t6tO cameru. and Sony YO-3800 sound re-
<br />corder. To prepare the video scenes for presentation to
<br />observers, the best (based on photographic quality and
<br />absence of unwanted intrusions such as auto noise) 12 s for
<br />each scene were selected from the 30 s recorded. Ifless than
<br />12 s of good quality footage was available, the scene was
<br />dropped. After eliminating poor quality video sequences,
<br />216 scenes remained, representing the 22 points at up to
<br />eight different flow levels.
<br />Row volume was estimated for each scene. Flows mea.
<br />sured al a U.S. Geological Survey (USGS) gauge down-
<br />stream of the inventory points were adjusted for diversions
<br />just upstream of the gauge, to arrive at the nominal flow
<br />levels listed in Table I. Inflows from tributaries and trans-
<br />basin diversions were then subtracted from the nominal How
<br />levels for points above these inflows to arrive at the final flow
<br />estimates for each scene.
<br />
<br />Scenic Beauty Judgments
<br />
<br />Scenes were presented to observers for their scenic
<br />beauty judgments in two formats designed to reflect two
<br />actual situations in which visitors might view the area. The
<br />rating format deemphasized flow rate as an important vari.
<br />able. It required each observer to rate the scenic beauty of a
<br />wide variety of scenes, differing in flow level and all other
<br />scene features. The mix of scenes is similar to what a visitor
<br />would observe during a summer on occasional drives
<br />through the entire length of river. Observers were informed
<br />that the purpose of the study was" to better understand how
<br />people perceive the scenic beauty of rivers and streams."
<br />that "some rivers have been specially designated for their
<br />scenic beauty," and that "public perception is an important
<br />consideration in selecting pans of rivers for special scenic
<br />designation." The wide range in topography, vegetation
<br />type, viewing perspective. etc., made it unlikely that anyone
<br />was aware of our primary interest in flow level.
<br />The paired.comparison fonnat forced some attention on
<br />flow rate by requiring observers to choose the more attrac-
<br />tive scene from among scene pairs, some of which depicted
<br />the same view at different ftow levels. This fonnat was
<br />designed to emphasize distinctions among flow levels, al-
<br />though as in the ratings format the instructions did not
<br />indicate a specific interest in any scene features. This format
<br />is more similar to the real world situation of returning to the
<br />same site or limited set of sites several times during the
<br />season. so that the visitor may tend to be more aware of how
<br />the sites look at different flow levels.
<br />Rating experiments. From among the 216 available
<br />scenes, 20 were randomly selected for the "baseline:' as
<br />required by the scenic beauty estimation (SBE) procedure
<br />llsed to scale the ratings [see Daniel and Boster, 1976; Brown
<br />and Daniel, 1990]. The remaining scenes were randomly
<br />Soned into three sets of 66 or 67 unique scenes each, and the
<br />20 baseline scenes were combined with each unique set for
<br />presentation to observer groups. A given set rarely included
<br />more than three scenes (i.e.. flow levels) of the same view.
<br />The 86 or 87 scenes in a set were randomly ordered, so
<br />that different scenes of Ihe same view were (ypically spread
<br />
<br />out among scenes of other views. The randomizing proce-
<br />dure was performed three times for each set. Presentation of
<br />the three random orders for each of three sets required a
<br />total of nine observer groups. Each group saw a separate
<br />specially prepared videotape. Randomization in assigning
<br />scenes to sets and in ordering scenes in a presentation help
<br />avoid order effects.
<br />The nine videotapes were shown in two experiments. one
<br />involving student observer groups at the University of
<br />Arizona and the other involving groups recruited from
<br />churches in the Fort Collins area. Previous studies of scenic
<br />beauty have shown close agreement between the judgments
<br />of university students and those of the general public [Brown
<br />and Daniel, 1986]. The Fort Collins groups were recruited
<br />because they were expected to be familiar with the Poudre
<br />River. In all, 85 Tucson observers and 113 Fort Collins
<br />observers rated video scenes, with at least 26 observers
<br />viewing anyone set.
<br />Presentation of the scenes to observers generally followed
<br />the procedure outlined by Brown and Daniel [1986). Observ-
<br />ers first listened to instructions read by the experimenter that
<br />explained the purpose of the scenic beauty judgment exer-
<br />cise and what was expected of the observers. The instruc-
<br />tions ex.plained that observers were to use a to-point scale,
<br />.. where a rating of I indicates low scenic beauty. and a to
<br />indicates high scenic beauty," and they were to use the
<br />photos to .. get an impression of what the represented stretch
<br />of river would be like if you were there:' Tucson respon-
<br />dents were informed that the areas "were filmed at several
<br />different times during the past summer" along "rivers and
<br />streams in the Rocky Mountains," while Fort Collins re-
<br />spondents were informed that the scenes were of the
<br />"Poudre River upstream of Fort Collins."
<br />Next, observers were shown 20 preview scenes that
<br />depicted the full range of types of scenes to later be rated, to
<br />allow observers to adjust their rating scale to the scenes at
<br />issue. Then the rating session began, with each scene shown
<br />for 12 s. Following the rating session. Fort Collins observers
<br />responded to a short questionnaire asking about their famil-
<br />iarity with and use of the study area.
<br />Observers' ratings within each of the three sets were
<br />adjusted for presentation order and then scaled using the
<br />program RMRATE [Bro,,'n er 01., 1990aJ 10 produce SBEs
<br />for each scene. Scaling to SHEs is a common procedure with
<br />rating data of scenic beauty and other environmental van.
<br />abies [e.g., Buhyoff er 01., 1982; Hull and Buhyoff, 1986:
<br />S{'hroederand Daniel, 1981]. The SHE transformation relies
<br />on Thurstone's law of categorical judgment [Torgerson,
<br />1958] and produces an interval-scale index. of ratings that is
<br />adjusted for the potential problems of simple rating scales,
<br />indicating relative perceived scenic beauty. As with any
<br />interval (nonratio) scale, the origin is arbitrary. The origin of
<br />the SHE scale is defined by the baseline scenes. The average
<br />baseline scene is assigned an SBE of O. Since the set of
<br />baseline scenes was randomly chosen for this study. the fact
<br />that a scene has a negative or positive SBE is unimportant by
<br />itself.
<br />Fon Collins observers were. quite familiar with the area;
<br />74% of them had lived in the Fort Collins area for 5 or more
<br />years, and 95% had been to [he study area. Fort Collins
<br />observers took a median of 4.5 trips per year to the area over
<br />the past 2 years. and nearly all of those trips were for
<br />recreational purposes. Fishing was the primary purpose of
<br />
|