SNTEMP (In)Frequently Asked Questions: Shade Issues

Q39. We have deciduous cottonwood trees on the banks and density measurements taken when the leaves were out. We will use these in one of the columns on the shade file. Is there a rule-of-thumb factor to apply to get a minimum shade value, i.e., do we multiply the summer value by .05, .10, … or what?

A39. I have never measured an un-vegetated deciduous riparian stand, but 5-10% sounds like a good guess. Perhaps you could take some measurements now, even around your own town, to get an idea. [Added 12/2001]

Q53. I am presently engaged in conducting the stream survey – unfortunately the stream is more like a river, varying between 400-100cfs. I have not yet been able to go full steam ahead on the modeling project, as the powers that be aren’t enamored of models. At his point, I am trying to collect a full set of data because I have been given the tentative approval for modeling in the future. Thus, I am gathering shade/canopy cover data. The use of a densiometer is precluded by the force and size of the river. I have been looking over the SRSHADE model as a substitute and notice that the data requirements are quite extensive. For width, I have 1:12000 aerial photo of the area, however, the stream is very braided. Is 1:12000 detailed enough? Do I sum the braided areas of the river and call it good?

A53. Though 1:12000 may be adequate, I believe you would get a better model if you ground-truthed you estimates with actual measurements, either with a range-finder or tape. For the temperature transport models (either SSTEMP or SNTEMP) it is the wetted width that is important, i.e., that surface in contact with the air. If your channel were only sparsely vegetated, I wouldn’t worry too much about shade. If more fully vegetated, it will take some guesstimates to arrive at a shade value that you feel comfortable with as there is no good way to composite the individual channel values that I know of. The Grey card technique might work well. [Added 12/2001]

Q54. Topographic Altitude: In looking at a topographic map, how do I estimate horizon?

A54. With great difficulty; unless you have a good feel for the location of the horizon, using topographic maps is a poor way to go. A GIS can do a tolerable job, but again I believe that selected ground truthing with a clinometer is by far the best approach. However, I’m wondering if topography is very important? Most streams with braided channel don’t have much effective topography. You may just be able to use a constant value of 5-10 degrees. [Added 12/2001]

Q55. Vegetation: How should I determine the outer limits for vegetation considered to have shade potential, given the width of the river and the height of the vegetation?

A55. Mostly by standing in the river and looking to see what can cast a shadow. [Added 12/2001]

Q56. Vegetation Crown: To estimate crown, do I need a more detailed aerial photo than 1:12000?

A56. As before, I believe that on-the-ground measurements, even if only a few, will provide a better characterization of crown than from aerial photography. Someone skilled in photo interpretation may be able to answer your question better than I. [Added 12/2001]

Q57. Is there an easier way to do this? Someone suggested using measurements taken alongside the bank with a densiometer as it was pointing in to the river – here we would need both sides of the river and average I suppose – however in some cases, only one side of the river is accessible.

A57. Maybe the best way is to ask yourself whether shade is worth measuring in detail for your river. Try putting together a crude SSTEMP model and see if shade is sensitive in the predictions for meteorological and hydrological conditions you wish to model. If it is, then you will need to get into the field and take some measurements. If it isn’t you could possibly calibrate using shade as a calibration parameter. If in-between, some good estimates may get you a long way without truly detailed measurements. [Added 12/2001]

Q63. I assume that Gray Card readings, if available, are use to represent vegetation density (Vd) in the model, i.e., a gray card reading that results in a Shade Quality of 0.5 equals a Vd of 50%?

A63. True. SSTEMP requires percent, whereas SNTEMP requires decimal fraction. [Added 12/2001]

Q64. Since gray card measurements are based on visible light, do you know of any "correction factors" for relating these gray card measurements to short-wave radiation?

A64. Good question. I have "assumed" that the two were highly linearly correlated. Let me know if this is not a good assumption. [Added 12/2001]

Q67. I am developing a Temperature TMDL using SNTEMP. If you use a shade file you give it east and west parameters, but I assume that this just produces a single shade value like is done in SSSHADE. But if you don’t use a shade file you have to give it two numbers (max and min). How does SNTEMP use the maximum and minimum shade values given it if the shade file is not used? Does it just take the mean of the two numbers, or some other combination?

A67. This is something that is not well explained in the manual. Basically SNTEMP will fit a sine wave through the max and min values in an attempt to simulate leaf-out and leaf-fall through the calendar year. I think it pegs the maximum value to June 21 or something like that. You can, and should, always double check the shade values listed in Table 6 (or is it table 3?) to make sure that the model is doing as you expect. [Added 12/2001]

Q93. Using the SSTEMP model, do you think there would be a way to estimate at what stream width riparian vegetation would be incapable of providing shade, even if it were old growth redwood at about 200 ft. height? Set topographic shade to zero, set latitude to 41 for Northern California, and set date to June 21 for highest solar angle. Set orientation to 0, and run it at 90 as well. In the SSSHADE module, what exactly does the vegetative shade value represent? Is it for an entire 24-hr. period? Is it over a given length of stream, or just in front of the trees that have the entered values for ht., crown diameter, and density? I do not see where in the SSSHADE module you set the segment length and time step. Perhaps that is done in another module.

A93. Vegetative shade simply relates to the portion of the sun’s energy intercepted by vegetation for the entire daylight (sunrise to sunset) period of one day. It assumes a homogeneous reach with the attributes you specify running the full length of the hypothetical segment essentially to infinity (at least this is the best way to think of it, I believe). [Added 12/2001]

Q95. Does the height of the shade above the stream make a difference in the temperature? Consider a small stream that is nearly 100% shaded by 10' shrubs, vs. the same stream shaded by 150' trees. Is it possible that the stream shaded by shrubs is warmer, because the solar radiation has warmed the surrounding air and created a warmer microclimate?

A95. I am unaware of any research on this issue, but speculation is certainly possible. Per SNTEMP/SSTEMP, shade is shade, so there would be no difference. BUT tall trees (implying a bit of a forest) likely reduce the ambient wind a bit more effectively than shorter, shrubby vegetation. Reducing the wind will influence evaporative heat flux. Depending on the exact type of canopy, it would also be possible that the microclimate may be more broadly impacted. This would include the efficiency of trapping and reradiating the water's back radiation (especially at night) but exactly how all of this would fall out is another question indeed.

Cindy Lowney with UC Davis Civil & Environmental Dept has been studying these sorts of issues. In particular, she's looking at how adjacent land use affects stream temperatures. Though this is not quite the nail on the head, she may be able to shed some light on this. I can't seem to find her phone number at the moment, but I'm sure you could get it from Gerald Orlob, 916-752-1424. [Added 12/2001]

Q148. I'd like to learn more about how sensitive the model might be with regard to shading, and if it could be useful in determining the placement of management activities (riparian revel) as would be required in a TMDL framework to attain/maintain some predetermined temp target.

A148. Yes, SNTEMP (and SSTEMP) are designed to deal with how shading affects stream temperature. Estimating the values to describe shading can be tedious and somewhat subjective, but it can certainly be done, especially if the goal is not 100% accurate results but rather evaluating the relative merit of different shading alternatives. I would recommend downloading SSTEMP and its documentation from http://www.fort.usgs.gov/ to get a feel for the data requirements and sensitivities. I think this would go a long way to answering your questions along this line. SNTEMP uses the same data, just in a more structured (and complicated) arrangement. [Added 6/2002]

Q149. Can you explain what's happening in the detailed shading algorithm? In particular, I've noticed that you get shading from "trees" that have height but no crown diameter. How is that possible?

A149. The riparian vegetation is like an unusual "L" shaped wall where the vertical stem is located with respect to the edge of the water at a distance of the offset. The vertical stem is only as solid as the its density; i.e., it can have holes in it that pass the sunlight just like the sun filtering through the leaves of a tree. The horizontal base extends out from the vertical stem a distance equal to one half the crown diameter, which is the radius, and is just above the elevation of the water surface. If the offset is equal to or greater than one-half the crown diameter, than the crown diameter has no effect on the shade.

Note that whatever part of the crown diameter extends over the water gives 100% shade (regardless of the Julian day, latitude, or azimuth of the stream) for that, and only that portion, of the water surface. For example, if the stream is 100 feet wide and there is riparian vegetation offset 10 feet on each side of the stream and the crown diameter is 30 feet, than the vegetation would extend 5 feet over the water surface on each side ([30/2]-10=5). That means that there will always be at least 10% shade (5+5=10) due to the overhang of the riparian vegetation. The vertical stem gives shade relative to the position of the sun (Julian day and time of day), stream direction (stream azimuth), and density of vegetation (sunlight filtering through the tree leaves).
[Fred Theurer] [Added 6/2002]

Q150a. I need to begin collecting some of the field-collected variables soon, but I'm still unsure about the collection methods for vegetation density.

A150a. My first, and really only question, to you is whether shading is an important management concern. If it is, then extra effort is warranted. If it is not, then I wouldn't get too concerned with details. One pretty good first step regardless is to do a quick and dirty sensitivity analysis using SSTEMP. (If you don't have the windows version, download it from our web site.) Play with the model (specifically density) to see if shading is really a sensitive variable. [Added 12/2001]

Q150b. I have a convex spherical densiometer that I had planned on using, but according to the IFIP No. 13 it seems this is not the best tool.

A150b. To tell you the truth, I'm not sure what the best tool is. I'd love to do a comprehensive comparative study. Lots of people use densiometers. What little experience I've had with them suggests to me that they are even more subjective than using a light meter, which has its own set of problems. [Added 12/2001]

Q150c. Is the vegetation density variable insensitive enough that using a densiometer would be okay?

A150c. See above

Q150d. I had thought about using the methods presented in Platts et al. (1987) for a concave densiometer and somehow reversing them for use with my convex densiometer (I assume you are familiar with the methods in Platts et al so I'm not going to detail them). From those methods I could obtain percent canopy density, and then from aerial photos I could obtain percent shade continuity for a reach; multiplying the two percentages would give me an average percent vegetative density for the reach (am I correct on this line of thinking?).

Q150e. If this method is okay I am still not sure about an appropriate amount of measurements to make for a reach of X length.

A150e. Always a slippery question. All I can say is, it depends. If you are really concerned, I think the Platts pub details taking a set of preliminary measurements, calculating the standard deviation, and then plugging into a formula that tells you how many additional samples you need to properly estimate the true mean. Frankly, I believe that you are better off subjectively judging the relative homogeneity of the streamside shading (don't forget the topographic shade) and just trying to take a few measurements to see what the values are. After a while, you begin to calibrate yourself, so to speak, and you will likely find that you become a pretty good judge. Keep taking measurements though, to keep yourself honest. [Added 12/2001]

Q150f. Nor am I sure about how to best separate density measurements for each side of the stream since Platts et al. methods take measurements from the stream banks and center.

A150f. I can't help you here as I don't know. Remember that the "best" measurements for SNTEMP would be taken from the center of the stream also. This applies primarily to topography (including the bank slopes), and less to vegetative shade that may need to be collected individually _if_ the vegetation differs from side to side -- an unlikely occurrence except in heavily modified landscapes. [Added 12/2001]

Q150g. Should measurements be made for each side of the stream?

Q150h. (I assume the model compensates if the stream does not run north-south with the azimuth variable) or for compass directions east and west?

A150h. Yes, it compensates. Make sure, however, that you update your manual with the errata because ip16 is wrong on east versus west. [Added 12/2001]

Q150i. Or is there a better methodology than presented by Platts et al to collect vegetation density over a stream with a densiometer?

A150i. I know of none. [Added 6/2002]

Q151. I have a quick question about how to go about using some of my data in SSTEMP and SNTEMP. I collected offset, crown, topo altitude, veg height, and shade quality at random locations within multiple reaches for both sides of the stream over the length of stream to be modeled. The data will then be averaged per reach.

If the location of a transect fell where there are no trees along side the stream, thus my offset, crown, and veg height would be nonexistent, should I average the variable for that transect as zero or just not at all? Also, if there were no trees for a transect should the 0% shade quality be averaged in as zero to account for shade continuity or not at all?

A151. Averaging randomly collected data, including all zeros, sounds good to me. Just make sure the answers seem reasonable when you are done. If you have enough samples, the shade calculations should work fairly well. [Added 6/2002]

Q152. I have a question about the max and min shade density variable needed in the shade file. I read the FAQ that addressed this, but I'm not sure how to get this variable out of my data. I measured shade density for both sides of the stream with a light meter and gray card at multiple locations within each representative reach during July. I then averaged these values to obtain a shade density for each stream-side for each reach.

1. Is the "maximum density for deciduous trees" the averages I have collected since they are for July? And the "minimum density for deciduous trees" a value from say December?

2. Is the "maximum density for deciduous trees" the largest density reading I measured for a reach and the "minimum density for deciduous trees" the smallest value measured?

Also, in the IF312 manual Table III.10 Shade File Format lists that veg crown measurement (field 9-16) is before veg height (field 17-24), but IFIP No.16 page III-92 says the opposite. Which is correct?

A152. I know this whole area is confusing. Unless you really need to simulate leaf-fall and leaf-out, I recommend NOT using the maximum value in the shade file. If you don't supply the maximum value, the model assumes that shade density is constant, which is usually sufficient. If you feel that you do need to simulate leaf-fall, I would recommend using a density of about 10% (less depending on the continuity factor) to account for trunk, branches, and stems of deciduous trees. As I understand it, the model fits a sin wave through these min and max values, but I'm not sure of the details. Make sure you look at Table 3 (or is it 6?) to see that the shade values actually used appear reasonable.

(Added later) If you believe that you can get by with a constant shading factor for each reach, do it. Data collection is at least conceptually simpler, and the whole affair makes more sense to most people. Unless shade is a legitimate management concern, it is not likely worth your while to collect the detail necessary to run the full shade model.

The field descriptions in IP16 are reversed. The errata at the beginning of the manual (depending on the version) should have corrected this. If for some reason you do not have an errata sheet, I believe you will find a file called ERRATA16 or something like that that came with your software. I recommend going through the errata and making all the changes in your manual. [Added 6/2002]

Q153. The shade file format described in Instream Flow Information Paper #16 (p. III-92) is different than that described in your self-study course (p 96). I'm wondering which is correct. Is vegetation height supposed to be in position 9-16 or 17-24?

A153. The self-study course notes are correct. Please see the ERRATA file that came with SNTEMP and update your IP16 accordingly. [Added 6/2002]

Q154. Is the vegetation density variable in the shade file entered as a percent or decimal in SNTEMP? No units are listed in the shade file format info. SSTEMP uses percent and I thought SNTEMP used a decimal.

A154. You are correct. SNTEMP uses a decimal fraction between zero and one. SSTEMP uses percentage. [Added 6/2002]

Q160. I have collected shade variables for a whole set of reaches and want to calculate shade for each one without using SSTEMP. Can your software be adapted for that?

A160. No, currently SSTEMP can only process a single reach for a single day. One user, Kent Smith, however, did implement the shading algorithm as an Excel spreadsheet. Though Kent does not maintain this spreadsheet, now can we officially endorse it, he has generously agreed for it to be distributed from our Web site. Download the software, SSSHADEX.XLS.

Q255. I’m working on a what-if scenario for an SNTEMP study. The stream is poorly shaded at present. If I put in a vegetation density of 70%, I am taking that to mean that a person standing in the stream and looking ashore would estimate that about 70% of the bank was vegetated. But it would not mean that 70% of the light would be intercepted by the vegetation. Do I have that right, or is there some other interpretation?

A255. From the notes: "Vegetation Density (percent) - This is the average screening factor (0 to 100%) of the shade producing strata of vegetation along the stream, Vd in the diagram. It is composed of two parts: the continuity of the vegetative coverage along the stream (quantity), and the percent of light filtered by the vegetation's leaves and trunks (quality).

For example, if there is vegetation along 25% of the stream and the average density of that coverage is 50%, the total vegetative density is 0.25 times 0.50, which equals 0.125, or 12.5%. The value entered into SNTEMP should always be between zero and one. Enter a value for both East and West Side."

You are correct. A person looking at the vegetation would estimate that, if the sun were shining through the vegetation to the stream that 30% of the radiation would get through. Then the model calculates how much of the time that vegetation is "effective" at producing shade.

Q256. I was reading through the SNTEMP FAQs regarding SHADE issues. I came across Q67 which is similar to what I would like to do. I have already run SSHADE and have an overall value that I'd like to use. But I can't find where the minimum and maximum shade values go that the questioner refers to. I thought they might go somewhere in the JOB file but nothing there.

A256. I am pleased that you've been looking through the FAQs, but sorry that Q67 was not as clear or thorough as it could have been. The minimum and maximum shade values would be entered in the stream geometry file. See page III-88 of Information Paper 16 and look for fields 41-48 and 49-56

Q257. In Information Paper 13, on page 26, you recommend that shade measurements be made only in two cases; otherwise, "windshield surveys" are satisfactory. Then pages 26-34 describe the measurement techniques. I just want to clarify that those ensuing techniques are only for the two special cases and do not constitute the "windshield survey." [For all I know, there maybe even more detailed ways of acquiring these data.]

The "windshield survey" does not involve any equipment, just observation. It gives a single value that can be use in SSTEMP in the "Total Shade" input box based on the observer's estimate of the percent of stream surface in shade all day.

A257. Basically, the issue revolves around whether shade is a potential management issue or whether it is just meant to represent the current (and relatively unchanging) system. If it is a management issue (not likely in Alaska, but maybe I don't really know), detailed shade measurements are absolutely required. If shade is simply a reach-by-reach 'parameter', then a windshield survey is fine. However, unless you are experienced with shade measurements, I still recommend taking some measurements to familiarize yourself with the techniques, and their variation (uncertainty). Usually, I find that after I take a few measurements (and I'll leave that number unquantified) I find that I can get reasonably good at estimating values for additional reaches. And then I also know approximately what level of variation I can use as calibration 'wiggle room' if required.

Q258. I'm currently trying to put together my input files for an SNTEMP model where I want to look at the effects of vegetation on stream temperature. I'm getting a little confused with the different data that is required as input when I want to use a shade file.

In the stream geometry file, two of the data fields are minimum stream shading and maximum stream shading (both in decimals). Reading the supporting model materials I think I have figured that these numbers are obtained from the % of solar radiation blocked by the riparian vegetation multiplied by the percent of each bank that has vegetation. I assume that the minimum and maximum allow for the reduction in shading abilities of deciduous trees during the fall and winter.

In the shade file there are fields for maximum and minimum vegetation densities on both banks of the stream. How are these numbers entered? Can I use the % of solar radiation blocked by the vegetation or am I supposed to use actually densities of the vegetation as in number of trees/hectare?

A258. Let me start by saying that you are basically right on the minimum and maximum values in the stream geometry file. Values in these locations are rarely used. In fact, I recommend that you not go that way unless you can get by with treating shade very simplistically. There is some relevant history here. Long ago when SNTEMP was being run on a mainframe computer, calculating shade was costly because shade was dynamically calculated for each reach and each time step and the trigonometric functions were time-consuming. For this reason, the developers set up the software so that you could calculate dynamic shade once, and then plug the values into the geometry file and skip calculating dynamic shade from then on. Fortunately, this is no longer necessary on today's PCs.

So, I recommend always using the shade data file. Now, what are all the values you need to fill in? First, I hope you have gone through the errata for Information Paper 16 and found that the manual is wrong on page III-92. Basically, the crown measurement comes before vegetation height, for both the east and west sides. You are correct that the vegetative density is the multiplicative combination of shade density and vegetative continuity. But why the maximum and minimums? The software does allow for leaf out and leaf fall, but in a rather crude way that might or might not work in your part of the world. Since most models are run just for the summer period when leaf-out is essentially constant, I recommend just filling in columns 33-40 so that SNTEMP will treat the shade as constant through time. You can experiment with this some more if you wish and then keep an eye on the shade predictions in the output files to see exactly what the model does.

Now I may not have adequately answered whether you can use the percent of solar radiation blocked by the vegetation or the actual density of the vegetation (as in number of trees/hectare). The number that SNTEMP "wants" is the mean area of the stream's surface without direct short wave solar radiation from sunrise to sunset. The way I think of this is like where we began, i.e., by multiplying the fractional density of an average tree by the continuity factor for those trees along the bank. It is far easier to make this parameter into something far more difficult than it really is. There is undoubtedly a lot of subjectivity here because the environment is (or at least can be) very heterogeneous. The more accurate you want these numbers, the more samples you need to take, and it can quickly become not very cost-effective.

Bottom line, unless you really need to do an extremely accurate job on shade, I recommend estimating the parameters well but quickly and then using shade as a calibration factor within "reasonable" bounds.

Q259. At a few of our locations, the angle of the embankment is greater than the angle to the topographic horizon. In fact, I am unable to see over the embankment while attempting to measure the topographic horizon. Should the embankment therefore be the measured angle? And in areas where the streambank is lower than the horizon and there is no vegetation, how should I take into account the height of the embankment?

In order to determine the average values of a segment (for vegetative height, offset, crown, etc), are measured values simply averaged or can the values be weighted to find the average? For example, if there is a segment dominated by one vegetation type with a smaller section of another vegetation (yet not significant enough to break up the segment into two), would the values be weighted due to the fraction of vegetation type to the total or would they be averaged equally?

This question is similar to another question regarding how to take into account areas of no vegetation. Much of our watershed is dominated by agricultural practices, therefore leaving only small sections of riparian communities (mostly seasonal crops and tall weeds are along the banks).

A259. The questions you have are common issues and mostly just take careful thought to deal with. What you wish to measure are those features that result in shading the stream, either by changing the time of local sunrise or sunset, or by filtering a percentage of the sunlight that would fall on the stream. If the stream bank dominates the "topography", then that becomes the topographic altitude that you would measure and vegetation would be set to zero. The case where vegetation "sits on top" of an elevated streambank is less straightforward. In these cases I usually recommend treating the whole "face" as you would treat vegetation, except the streambank would be 100% dense and constitute a fraction of the overall shade density factor. For example, let's say that looking at the entire "face" in front of you, perhaps 66% is bank and 33% is vegetative shade. Assuming the vegetation is 75% effective shade, the entire field of view would then have a "vegetative" density of:
0.66*1 + 0.33*.75 = 0.90

Analogous to the weighting I just outlined, it is certainly appropriate to weight portions of a stream segment to reflect any significant changes on a reach-by-reach basis.

Q260. To follow up on the question about stream bank as the topographic altitude, what if the stream bank is only 2-3 feet high and my altitude readings are taken from eye level so the bank doesn't influence the altitude I'm measuring - would I still take the bank into account? Also, I suspect I need to take into account my eye level when determining the final number for the topographic altitude, but how do I do this? Finally, if it's impossible to determine the altitude from the stream (due to thick vegetation), can we just use a topographic map to calculate the angles (and wouldn't this angle is some cases actually be more exact)?

And do you have any tips for measuring percentage vegetation cover along the stream? I'm hoping to use both an Orthophotoquad as well as field measurements and I was just wondering if there was a standard method to estimating these percentages.

A260. The honest truth, though I am sometimes loathe to admit it, is that we can all go to a lot of trouble to make our shade estimates "exact", but they will still be just estimates. Yes, ideally one would 'look' from the water's surface. Frankly, I rarely bother. If the bank is so low that you can see over it, then it probably doesn't restrict the local sunrise/sunset by more than a few degrees, likely within the realm of sampling error anyway. Unless your study (and I suppose I should have asked) is explicitly to evaluate shading as a management objective, it is probably not worth going to a lot of trouble. If shade is an objective, then some simple geometric calculations (or bending over occasionally to get an idea of the angular difference) might pay off if you are careful

I'm not trying to make light of the issue, but I believe we must remember that shade estimates have considerable uncertainty. Some of that uncertainty is troublesome in terms of evaluating the 'goodness' of your model, but some of that very same uncertainty is a blessing in terms of providing a rationale for model calibration.

Many folks today are using relatively sophisticated GIS software to calculate overall topographic angles. Unless you have access to cheap services, I suspect that this is far too costly for most applications and results in a product that is viewed as highly accurate (it isn't) and useful (it may not be because of stream banks, etc.) I have recently begun to use a simple technique using the state-by-state MapTech product (available over the web for about $100). [Please note that USGS does not endorse commercial products.] It is somewhat laborious, and not straightforward, but you can do a fair job of estimating the topographic horizon and estimating the distances horizontally and vertically to calculate the topographic angle if it is a ridgeline and not the streambank. The topographic maps are also OK for estimating vegetative coverage if the underlying USGS map is fairly new and not much has transpired in terms of land use change.

As far as measuring percent vegetative shade, I still like the gray card technique discussed in Information Paper 13, though there are many competing methodologies out there. Whatever technique you use, I believe you will find it relatively straightforward after a few measurements to characterize the general shading from various vegetative complexes. I find a great deal of homogeneity, or perhaps rather repetitiveness. Then, as far as vegetative continuity, I generally just do a "windshield" estimate. But again, if shade is critical, I would recommend taking 10 or so samples per reach and averaging them. This at least would seem more accurate, even if there is still considerable room for judgment, and you will get fast at it after a while.

Q261. Hello my name is C. W. and I am a PhD student at the University of X. I have been working on the impacts of forest fire and debris flows on stream temperature in collaboration with the Rocky Mountain Research Station. Debris flows and, to a lesser extent, wildfires decrease the canopy over the stream. While this opens up the stream to increased shortwave solar radiation, it also appears to alter the long wave radiation balance. I looked up your 2000 paper in Rivers 7(4), as the removal of vegetative canopy from logging may be similar to canopy removal by debris flow. In your paper you do not discuss the alterations to long wave radiation balance in your modeling. But in looking at the equations Theurer used in developing SNTEMP, it appears that changing the shading of a stream, all else being equal, will change the import of long wave radiation to the stream. Using values for my study streams, it is possible to go from net import of long wave radiation in very shaded reaches and net export in very open reaches. I was wondering if this is an effect you have noticed in your work. I have found very few studies that discuss the impact of long wave radiation, and when they do, it is often dismissed as balancing out over a 24 hour period (Beschta, 1997 Rangelands 19(2)). Or they assume that shading has no impact on the Long wave budget (Johnson, 2004 CJFAS 61). Neither of these answers is in agreement with the equations used in SNTEMP, and I was wondering if this is an issue which has confronted you or your group.

A261. Thank you for your very insightful question. I wish I could give you a "Theurer" answer, but I am not intimately familiar with his equations, though I do use the model quite a lot.

I looked in Information Paper 16 and see what you are likely referring to in equations II(61), II(65) and the accompanying notes. As stated in the text, if atmospheric shade is indeed equivalent to solar shade, which is in turn related directly to vegetative shade, then I believe you are correct. I looked in the source code, and indeed found that the calculated shade was used in the subsequent calculation of both atmospheric (Ha) and solar (Hs) radiation.

Now, what is correct -- or more correct? I myself have never attempted to measure these variables, but Sherri Johnson has. (Have you contacted her to see what she says? I have her e-mail address if you do not. Let me know.) It seems to me that the real world quickly becomes complicated. In SNTEMP, shade can arise from either vegetative or topographic means. It would certainly seem to me that true ambient air temperatures would circulate more readily in the absence of dense vegetation, but in canyons, air might circulate less but retain heat longer for many reasons. I am not sure.

Let me add that SNTEMP "wants" as input your best estimate of air temperature. I always recommend that people use air temperatures collected in close proximity to the water's surface because that's where heat flux is taking place. Unfortunately, these data are rarely available, so people use meteorological stations on dams, at airports, etc., and modeling suffers somewhat because of it. And it also begs the question of whether air temperatures reflect shaded or unshaded stream reaches.

I'd be interested in your detailed perspective and how you believe modeling should best proceed.

Q262. Do you know of an available software program that calculates the topographic altitude angle for a stream location using digital elevation models (I have 10 meter Digital Elevation Models and LIDAR data)? I am aware of the TTool program that is part of the Heat Source model but TTools calculates the maximum topographic shade angle east, south, and west of a stream segment node when I need the angle perpendicular to the general stream reach azimuth for the SNTEMP model.

A262. I do not know of one, but I wouldn't think it would be too difficult to locate one from some of the GIS wizards.

Q263. First, are the units for “Eastside vegetation density (conifer); minimum density for deciduous trees” (record 5, Field 33-40 on page III-92) percent expressed as a decimal (1.00) or actual percent (100.0). I assume it’s the former but wanted to check.

Second, I added C nodes with shade data to the stream geometry file (GEOMDAT) and the shade file (SHADE). To run the shade routine in SNTEMP do you just need to set the flag in the JOB file (record 3 column 78) to “T” or are there other flags in other files that need to be set?

A263. You are correct on the percentage. I hope I got this right in the IF 312 notebook. You might check there. I also hope you looked at the errata because there is an important change from Theurer et al.

On number 2, I believe you must set both 78 and 79 to T. 78 says that data are present and 79 says to use them.