Dear GLIMS colleagues and over interested researchers/students: This message is (1) An announcement of the GLIMS dinner (TUESDAY Dec.10, Henry's Hunan Restaurant on Natoma St, San Francisco-- please go to the right one),and (2) a response to Graham Cogley's message on GLIMS 2 and the RGI. (1) GLIMS DinnerAt least for the GLIMS dinner, students and others who have not been a part of GLIMS are welcome. If we have more than 30 people expected to attend, I need to know soon to reserve a different space. Right now, I have a full private room reserved for 30 people at Henry’s Hunan Restaurant. RSVP (email me). I will confirm by 1 Dec. that either this is the venue, or a different venue,so please RSVP. The room absolutely will not take in more than 32 people. We will order communally and pay communally. We’ll be sure to have plenty of vegetarian dishes as well as others. Feel free to bring a significant other. TUESDAY (not Wednesday as Graham thought). 6:30 for the RGI discussion, and 7:00-7:30 for broader GLIMS discussions, and 7:30-9:00 for dinner. SOME DISCUSSION TOPICS ARE DESCRIBED BELOW. Henry's Hunan Restaurantmore info 110 Natoma StreetSan Francisco, CA 94105(415) 546-4999 (2) Comments by Jeff Kargel on GLIMS2 (please refer to Graham's email and attachment): I agree heartily with Graham’s comments overall, particularly his general call for assessment of a “GLIMS2” that will include shoehorning in the RGI but has to go beyond that. I agree also with his emphasis on multitemporal coverage. However, in dealing with multitemporal coverage we have to consider how such data will be used for differential analysis, first-order derivatives and second-order. For something like ASTER-to-ASTER, the problems (including error analysis) are relatively straight forward to do first-order change assessment, such as Cosi-Corr analysis of flow speeds or manual or semiautomatic assessment of glacier frontal retreat or advance, and a little more challenging (or impossible in many cases) to get ASTER-based elevation changes. (A careful reading of what’s in the GLIMS book assures the reader that even all that is rich with challenges.) A further level of challenge is in assessment of acceleration or deceleration of glacier changes. There is a lot of interest in this matter, and we should pursue it, but obviously the greater importance has to be attached to completing the single-snapshot global coverage for GLIMS/RGI and then simply the average rate of change over a given satellite-era time period. However, to deal with the first and second derivatives of change, we need a new focus on error analysis , because obviously errors mount the more one deals with the higher order derivatives or the shorter timespans. Even more problematic is when completely different types of datasets are used, for example, satellite multispectral vs. historic map archives. WE WANT TO INCLUDE/EXPAND IN THE GLIMS DATABASE MULTITEMPORAL DATA ON GLACIER STATE AND DYNAMICS. So I am not discouraging that; to the contrary. However, we really need to recognize—as a community—that the “garbage in, garbage out” mantra applies ever more strongly when differentials and especially higher order derivatives are concerned. So how are we to determine what is garbage in a given context for a given type of analysis? We as a community have—thankfully—embraced the idea of using historic data and diverse satellite data types. And overall, we have been careful to assess whether certain datasets can be used effectively with other types of data to obtain change information that otherwise cannot be determined. Rigorous error assessments can be much harder to determine when very different types of data are used, one of which we understand its error characteristics (such as ASTER) and another where we might not (some historic maps, for instance). Another big theme that we should delve into more deeply: we need a much more extensive set of weather station data accessible to the public science community, and we need to get beyond reliance on airport-based weather stations. We need data—even a single full annual cycle’s data—at the toe, near the ELA, and above the ELA of many, many glaciers, many hundreds of them. We need adiabatic lapse rates that can be believed (averaged over the annual cycle and seasonally/monthly resolved averages), and other information that are strictly pertinent to glaciers. This would help us apply standard national weather station data and global/downscaled climate models with greater confidence that we are discerning orographic influences and the actual vertical structure of the atmosphere. Full AWS data are great, but I think a lot can be done with simpler data recorders, but we need a standardized protocol. Furthermore, sensors such as MODIS and even ASTER provide a lot of information. We need better down-scaled climate models that can peer into cirque basins, and this work should move beyond the journal-article mode to a GLIMS-type mode of results dissemination. We need to make it more glacier-relevant and publicly available/user-friendly. In sum, we need to delve more deeply into the issue of error assessment for different types of data and different types of data analysis, and coordinated glacier-climate studies (among other things Graham mentioned). I propose that we form two working groups that will function a bit similarly to the RGI group: GLIMS Error working group, and GLIMS Climate working group. We have a strong basis to continue this discussion (all the excellent work that the GLIMS and RGI community has done, the GLIMS book as one example), and we can pursue this discussion at the RGI/GLIMS dinner. However, I think we need a full workshop on these. So here I propose, and we can discuss in more detail at the dinner, that we hold a next workshop on the topic of “Error Assessment and Climate Modeling Applied to Analysis of Changing Glaciers.” We can discuss the topic, the location, and the date. But to start that process, I suggest Stockholm, Sweden (workshop) and optional fieldtrip (strongly encouraged) at Tarfala Station. Other options may be Alaska, or Seattle, with potential for glacier fieldtrips perhaps being logistically simpler and cheaper than Tarfala Station. There are some other matters I want to bring up at the GLIMS/RGI dinner. Jeffrey S. KargelDepartment of Hydrology & Water ResourcesUniversity of ArizonaTucson, AZ 85742 USAEmail(1) jeffreyskargel@hotmail.comEmail(2) kargel@hwr.arizona.eduMobile phone: 520-780-7759www.glims.org