From uniola@email.unc.edu Wed Mar 19 14:53:25 2003 Date: Wed, 19 Mar 2003 13:38:16 -0500 (EST) From: Robert K. Peet To: Don Faber-Langendoen Cc: David Tart , Michael Jennings , Michael Lee Subject: RE: Picklists -- soils Regards VegBank, I see the cover issue as a metadata issue. To be rigorous, perhaps we should have a field specifically to record this, either in coverMehtod or in observation. My preference woulbe to put it in observation. MIke Lee, would you see that this enters Bugzilla? Bob On Wed, 19 Mar 2003, Don Faber-Langendoen wrote: > I'm content to add those missing classes so we can make the system usable to > both sets of users. Dave's point about trying to make VegBank FGDC > compliant on non-veg issues is important. Sometimes we tend to do things > because we think this will lead to "buy-in." But the issue of being FGDC > compliant on all non-veg issues is really beyond the scope of VegBank. We > do need to ensure that the databases can talk to each other. If the FS > creates a database structure that can't handle both a 12 point scale as well > as a 21 point scale, they've put themselves into a narrow framework that > works for them, but not for all users. We shouldn't do the same. > > I believe most ecologists use canopy cover estimates rather than foliar > cover estimates (but I have been with ecologists who do make minor > corrections for very open canopies). As Bob points out, these issues will > not impact the quality of the data for classification purposes. There's a > higher degree of real ecological variability that makes distinctions between > foliar and canopy cover estimates less than critical. Perhaps droughtier > ecological systems are more affected by the choice?. As long as we specify > which one is being used, I think we'll be fine. > > Bob, is this strictly a meta-data issue about cover, or are you considering > a field that needs to be added? > > Don > > -----Original Message----- > From: Robert K. Peet [mailto:uniola@email.unc.edu] > Sent: Tuesday, March 18, 2003 7:29 PM > To: David Tart > Cc: Don Faber-Langendoen; Michael Jennings; Michael Lee > Subject: RE: Picklists -- soils > > > > > Dave, > > With VegBank we are simply trying to retain sufficient flexibility to > support the needs of a rather broad user base. Thus, if FS is going to > routinely use a 21-point texture class scale and NS is going to routinely > use a 12 point scale, we need to find a way to accommodate both approaches > with minimal complexity for the users. > > How do we do this? I propose we accept the 21-point FGDC list and that we > add to it the missing classes of "Sands, loamy sands and sandy loams. I > also suggest that for purposes of our database we indict the subclasses by > having the name in the list read as follows Sands: sand Sands: coarse sand > Sands: fine sand Sands: very fine sand etc. This will avoid mistakes > resulting from those who want to enter "sand" and who have never noticed > the difference between "sand" sensu the NRCU Soil Manual soil triangle > (fig 3-16) and "sand" sensu NRCS SOil Manual Chpt 3 text and the FGDC > standard (which is a subset of the other sand. At the same time, it will > allow unambigous mapping to the FGDC standard. > > We don't view Vegbank as the standard. However, where we are headed is > toward a data exchange standard for plot data, rather similar in character > to XML. This will likely draw heavily from the VegBank ERD to develop an > appropriate logical structure. We anticipate this as a component of the > Veg Standard to be added to the draft as an appendix in perhaps 2-3 > months. At that point we will need some of the FS database guys to look > it over to see if it will work for them. We also have to recognize that > part of our data exchange specification will consist of extensions to > nonveg stuff like soil texture. We should probably include these but > identify them as extensions that can or should be replaced by data > exchange standards to be published by the other subcommittees. > > The discussion over the cover measures suggests to me that we are making > our standards a bit too tight. If we make a veg type sink or swim based > on use of canopy cover vs foliar cover, we have made some serious mistakes > in constructing the standards. We need to accept legacy data that are > somewhat divergent from best practice but which inform the process. We > perhaps should state in the document that best practice is "canopy cover", > but acceptable is "foliar cover". Good plot data from some regions are so > sparse that we need to accept and combine multiple types of data. > > Best, > Bob > > > On Tue, 18 Mar 2003, David Tart wrote: > > > > > All, > > > > To do as Don suggests, you will have to add Sands, Loamy Sands, and Sandy > > Loams to the picklist in order to keep these classes distinct from the > > subclasses sand, loamy sand, and sandy loam. However, using the sandy > > loams class would not follow Soil Survey Manual direction and therefore > > would not be FGDC compliant. So, 12 categories does not appear to be a > > viable option, while 15 is acceptable (currently). I suspect that NRCS > and > > Forest Service will reject anything other than the 21 subclasses, since > > this appears to be current practice in soils and is the only option > > supported by NASIS. > > > > It is likely that in the long run, only the minimum required data in the > > ESA document will end up in the FGDC vegetation classification standard. > > Equating VegBank and its picklists with the proposed vegetation > > classification standard is likely to hurt chances of federal acceptance of > > the vegetation standard. Given the possible ramifications of the Data > > Quality Act, the content of the FGDC vegetation standard will likely have > > to be limited to vegetation attributes only. Any conflict between FGDC > > standards will lead to lawsuits for federal agencies. I would hope that > > the Forest Service and BLM would prefer to avoid setting themselves up, by > > not allowing the FGDC Vegetation Standard to conflict with other FGDC > > standards. > > > > The Data Quality Act also makes the canopy vs foliar cover issue critical > > to the Forest Service. A standard that does not include canopy cover > would > > instantly brand FS legacy data as inferior or substandard. Then any FS > > management decision that could be tied back to vegetation types could be > > challenged on that basis. This is one example of the kind of issues that > > will come up in the FGDC review of the standards document. > > > > Dave > > > > David L. Tart - Regional Vegetation Ecologist > > Intermountain Region > > Vegetation Management (Room 5022) > > U.S. Forest Service, 324 25th Street, Ogden UT 84401 > > Office: (801) 625-5817 Fax: (801) 625-5483 > > > > > > > > "Don Faber-Langendoen" > > Tart" , "Robert K. Peet" > > eserve.org> cc: > "Michael Jennings" , "Michael Lee" > > > > > 03/18/2003 03:42 PM Subject: RE: > Picklists -- soils > > Please respond to > > don_faber-langendoen > > > > > > > > > > > > > > If we take the approach of including the 21, then we clearly need to > > indicate what the 12 basic ones from the triangle are in the description > > field, so that users can clearly follow the picklist. > > > > Don > > > > -----Original Message----- > > From: David Tart [mailto:dtart@fs.fed.us] > > Sent: Tuesday, March 18, 2003 5:18 PM > > To: Robert K. Peet > > Cc: Don Faber-Langendoen; Michael Jennings; Michael Lee > > Subject: RE: Picklists -- soils > > > > > > > > Bob, > > > > I did find a bit of a caveat regarding soil textures in the soil survey > > manual. Following the definitions of the classes and subclasses it says, > > > > "The eight distinctions in the sand and loamy sand groups provide > > refinement greater than > > can be consistently determined by field techniques. Only those > > distinctions that are > > significant to use and management and that can be consistently made > > in the field should be > > applied." > > > > > > > > > > This allows the option of using "sands" or "loamy sands" in describing a > > soil pedon. However, this does not appear to be the current proctice > > because the NASIS database does not have distinct codes for the sand > > subclass and the sands class, etc. (NRCS is in the same situation as FS, > > their protocol and database are out of sync.) Applying this rarely used > > rule would result in 15 texture categories instead of 21. (The rule does > > not apply to sandy loams.) The Soil Survey Manual is being revised, so > it > > is possible that this obscure rule may be removed. > > > > Dave > > > > David L. Tart - Regional Vegetation Ecologist > > Intermountain Region > > Vegetation Management (Room 5022) > > U.S. Forest Service, 324 25th Street, Ogden UT 84401 > > Office: (801) 625-5817 Fax: (801) 625-5483 > > > > > > > > "Robert K. Peet" > > > , Don Faber-Langendoen > > c.edu> > > , Michael Jennings > > , Michael > > Lee > > 03/18/2003 12:36 cc: > > PM Subject: RE: Picklists -- > > soils > > > > > > > > > > > > > > > > > > We seem to be at an impass regards the soil texture list. Let me try to > > state where I think we are. > > > > Let's agree that we drop all categories except all or a subset of the 21 > > types in the FGDC sils list. The crux is that much data is collected using > > the 12 categories on the traditional soil triangle. Other data are > > collected using all 21 categories in the FGDC list. I think I was > > initially confused and thought the 21 categories were mutually exclusive, > > but now I see the long list as having overlapping categories. For example > > coarse sand, fine sand and very fine sand nest inside sand. This pattern > > is repeated for sandy loams and sandy loans. Thus we do not have to worry > > about there being two meanings of sand where one is a subset of the other. > > Given this, there is no cost to retaining all 21 categories. If you > > happen to want to use the basic 12, great. If you want to use the finer > > scale cntaining 18 distinct types and three higher level types, that too > > is great in that the fine-scale cane be lumped up to match the coarse > > scale. > > > > IN short, we go back to 21 items in the closed list in full knowldege that > > they are not mutually exclusive. > > > > OK > > Bob > > > > > > > > > > On Tue, 18 Mar 2003, David Tart wrote: > > > > > > > > All, > > > > > > I recommend deleting the rock size terms from the list. I have attached > > a > > > copy of the list with the terms to be deleted in red. These terms are > > > applicable only in very rare situations where no soil matrix is present. > > > The soil scientists I talked to were not familiar with any examples of > > > this. > > > > > > The 21 textural classes for mineral soils are standard for describing > > soil > > > horizions. They are necessary for calculating Available Water Capacity, > > > which is one of the most useful soil properties relating to vegetation > in > > > the west. If an ecologist cannot identify the 21 textural classes in > the > > > field, then a soil scientist should be used to do the work. I would > also > > > question an ecologists ability to consistently and correctly identify > the > > > soil horizons themselves. > > > > > > More generalized categories are available for describing soils in a > broad > > > brush way, but these are not used to describe individual soil horizons. > > > The 12 classes shown on the textural triangle are the first level of > > > generalization above the 21 classes in the picklist. These can be > > further > > > grouped into either 3 or 5 categories. If you want more details I can > > > provide them. > > > > > > Dave > > > > > > (See attached file: VegBank Soil Picklist.doc) > > > > > > David L. Tart - Regional Vegetation Ecologist > > > Intermountain Region > > > Vegetation Management (Room 5022) > > > U.S. Forest Service, 324 25th Street, Ogden UT 84401 > > > Office: (801) 625-5817 Fax: (801) 625-5483 > > > > > > > > > > > > "Don Faber-Langendoen" > > > > "Robert K. Peet" , "Michael Lee" > > > eserve.org> > > , "Michael Jennings" , > "Gabriel > > Farrell" > > > > > , "David Tart/R6/USDAFS" > > > 03/17/2003 09:50 AM cc: > > > Please respond to Subject: RE: > > Picklists > > > don_faber-langendoen > > > > > > > > > > > > > > > > > > > > > Bob, > > > > > > I have added my comments below. One item you didn't list as changed was > > > soil texture. I strongly object to th proposed FGDC texture options! > > Few, > > > if any, ecologists collect soil texture to this level of detail. We > > either > > > need to make this hierarchical, allowing for multiple levels of > > resolution, > > > or go back to the more basic 12 categories that have some hope of being > > > keyed out in the field. We can make our picklists compatible with FGDC > > > lists without adopting them verbatim. > > > > > > Don > > > > > > -----Original Message----- > > > From: Robert K. Peet [mailto:uniola@email.unc.edu] > > > Sent: Sunday, March 16, 2003 1:48 PM > > > To: Don Faber-Langendoen; Michael Lee; Michael Jennings; Gabriel > > > Farrell; David Tart/R6/USDAFS > > > Subject: Picklists > > > > > > > > > > > > I have gone through the list of picklists prepared by Mike and made a > > > bunch of changes. I attach it for your consideration. Here are a few > > > highlights > > > > > > 1) commStatus.commLevel in generally used to house the hierarchy of the > > > NVC, much like we place terms like order, family, genus, subgenus, > > > species, variety here in the plant module. However, there is a problem > > > that different vegetation classifications have different hierarchies. > > > This suggests to me that at least for the vegetation classification we > > > need to keep this an open list, prepopulated with FGDC. Otherwise we > > > cannot accommodate the BB folks. > > > > > > **DFL. I'm not sure I understand enough about VegBank to follow this, > but > > > why wouldn't you maintain a separate set of lists for CommStatus for > > > different classifications. i.e. if you pick USNVC you get one list, if > > you > > > choose B-B you get another (if such a B-B list can be achieved). And > your > > > citation of B-B nomenclatural reference as a source isn't the same kind > > of > > > reference, since that citation lacks any syntaxonomic units per se. > > Seems > > > like our species taxonomic model would be more appropriate, i.e. you > > > specifiy an authority and you pick from the list?? > > > > > > 2) I have reworked commConvergence and plantConvergence > > > > > > 3) I have added a few types to disturbanceTYpe > > > > > > 4) I point out that namedPlace MUAST refer to four closed lists that we > > > have created but which are not in this sequence of closed lists > > > > > > 5) geology. First, I think this name is confusing everyone and we > should > > > change to rockType. I take Dave Tart's suggestion of matching FGDC > > > seriously here and have retained the list from the soil standard, for > the > > > present. I know that Don has put a lot of hard work into hte Geology > > > types, but I remain relativley happy with the Soils FGDC list I > suggested > > > a month ago and feel it captures some important difference missed in > > Don's > > > effort where acidic and basic types are not separated. > > > **DFL. I can live with this, though presumably this needs to continue to > > be > > > an open list, as the categories are not mutually exclusive. Is a user > > only > > > allowed to pick one item from the list. > > > > > > 5) Surficial Deposits. This is a new field that has been emerging over > > > the last month. Note that once again I expunge the word geology. This > > > time I find I like Don's revision of my revision of ... I have, > however, > > > seen fit to add a bunch of categories. The current list is as follows: > > > **DFL, several changes recommended here: > > > > > > Add : Glacial Deposits: Glaciolacustrine > > > Add: Glacial Deposits: Glaciomarine > > > Reword: Glacial Deposits:Glaciofluvial > > > Split: all three Marine and Lacustrine Deposits into separate Marine > > versus > > > Lacustrine. Thus: > > > Marine Deposits: Unconsolidated Sediments > > > Marine Deposits: Coarse sediments > > > Marine Deposits: Fine-grained sediments > > > Lacustrine Deposits: Unconsolidated Sediments > > > Lacustrine Deposits: Coarse sediments > > > Lacustrine Deposits: Fine-grained sediments > > > > > > ****************** > > > Residual Material: Bedrock > > > Residual Material: Disintegrated Rock > > > Residual Material: Deeply Weathered Rock > > > Glacial Deposits: Undifferentiated glacial deposit > > > Glacial Deposits: Till > > > Glacial Deposits: Moraine > > > Glacial Deposits: Bedrock and till > > > Glacial Deposits: Glacial-fluvial deposits (outwash) > > > Glacial Deposits: Deltaic deposits > > > Alluvial Deposits: Floodplain > > > Alluvial Deposits: Alluvial Fan > > > Alluvial Deposits: Deltas > > > Marine and Lacustrine Deposits: Unconsolidated Sediments > > > Marine and Lacustrine Deposits: Coarse sediments > > > Marine and Lacustrine Deposits: Fine-grained sediments > > > Organic Deposits: Peat > > > Organic Deposits: Muck > > > Slope & Modified Deposits: Talus and scree slopes > > > Slope & Modified Deposits: Colluvial > > > Slope & Modified Deposits: Solifluction, landslide > > > Aeolian Deposits: Dunes > > > Aeolian Deposits: Aeolian sand flats and cover sands > > > Aeolian Deposits: Loess deposits > > > Aeolian Deposits: Volcanic Ash > > > Chemical Deposits: Evaporites and Precipitates > > > Other > > > Variable > > > > > > > > > I suppose the rockType list is now the most problematic. However, I am > > > going to stick with the FGDC list until we generally agree to do > > > otherwise. > > > **DFL. This rockType list will require remapping of PLOTS to VegBank. > > > > > > Once we agree on this, we need to select which subset of these lists we > > > want to include as part of the ESA Guidelines version 2 (Appendix). It > > is > > > now evident that a lot of FGDC standards do embed picklists. > > > > > > **DFL. I have the lifeform table complete, but need to double-check. > > I'll > > > send it out later thisweek. > > > > > > Best, > > > Bob > > > > > > > > > > > > > > > On Mon, 3 Mar 2003, Don Faber-Langendoen wrote: > > > > > > > Bob and Mike, > > > > > > > > Here is my revised picklist for geology. I received comments from a > > > number > > > > of NatureServe ecologists that helped to clean it up. But I still > > don't > > > > know if Dave Roberts thinks that this works for the West. In any > case, > > > here > > > > it is. We will need to redo the loader for the NPS PLOTS database to > > > match > > > > this new format, though it shouldn't be hard. > > > > > > > > Mike, I still have that lifeform/physiognomic list to work on, but I'm > > > not > > > > sure what I was asked to do on that. I have spent a fair bit of time > > > > cleaning up the structure of our Biotics database on this same issue, > > so > > > I'm > > > > hoping that what I've done there will translate to VegBank. > > > > > > > > Don > > > >