1. USFS Region 1 SAS Analysis of Lake Chemistry, NADP, and IMPROVE data Jill Grenon and Mark Story, Gallatin NF R1 Air Quality Monitoring Program Overview Non Parametric Statistical Methods Statistical Test Results Tentative Conclusions

2. USFS R1 Wilderness Air Quality Monitoring Plan Mark Story - Gallatin NF, Thomas Dzomba - USFS R1/R4 AFA, Jill Grenon - Gallatin NF/MSU 2/1/2008 INTRODUCTION Protection of air quality values is a key component of both the Clean Air Act and Wilderness Act. The USFS Region 1 has 13 designated Wilderness areas. In terms of air quality, seven are designated as Class I Wilderness areas and six are designated as Class II Wilderness Areas. Class l areas in USFS R1 were designated by the Clean Air Act amendments of 1977. The 1977 Clean Air Act amendments assigned the Forest Service an “affirmative responsibility” to protect the Air Quality Related Values (AQRV’s) of Class l areas. Class II areas include all other areas of the country that are not Class I. Class II Wilderness areas are Class II for the Clean Air Act Prevention of Significant Deterioration (PSD) regulations. Air quality protection authority (beyond ambient air quality standards and PSD increments) for Class II Wilderness areas therefore relies primarily upon the Wilderness Act with the air quality values titled Wilderness Air Quality Values (WAQV’s). Region 1 has been actively monitoring AQRVs and WAQVs since 1989. Formal AQRV monitoring plans for regional Class I Wilderness areas were developed between 1989 and 1996. For Class II Wilderness areas, formal WAQV plans were developed in 2007 and 2008 in accordance with the 10-Year Wilderness Challenge. The following table summarizes the plan development for each Wilderness area; each plan is referenced at the end of this plan and tabulated below. http://www.fs.fed.us/r1/gallatin/resources/air/

3. Wilderness AreaClassAQRV or WAQV AQRV or WAQV values Bob Marshall (BMW) 1AQRVvisibility, aquatic ecosystems, wildlife Cabinet Mountains (CMW) 1AQRVscenery, aquatic ecosystems, vegetation, wildlife Gates of the Mountains (GMW) 1AQRVvisibility, water, wildlife, flora Selway-Bitterroot (SBW) 1AQRVvisibility, aquatic ecosystems, soils and geology Anaconda-Pintler (APW) 1AQRVscenery and visibility, water quality, wildlife, vegetation, fragrance, wilderness experience Scapegoat (SGW)1AQRVvisibility and scenery, water quality, wildlife, vegetation, odor, climate Mission Mountains (MMW) 1WAQVvisibility and scenery, aquatic ecosystems, vegetation, wildlife Absaroka- Beartooth (ABW) 2WAQVvisibility and scenery, alpine ecosystems, wildlife Lee Metcalf (LMW)2WAQVvisibility and scenery, lakes, wildlife Great Bear (GBW)2WAQVvisibility and scenery Rattlesnake (RW)2WAQVvisibility and scenery, lakes Welcome Creek (WCW) 2WAQVvisibility and scenery Gospel Hump (GHW) 2WAQVvisibility and scenery, lakes

4. monitoring item BMWSGWCMWSBWAPWGMWMMW Phase 3 Lakes Upper & Lower Libby Lakes North Kootenai & Shasta Lakes IMPROVE visibility MONT1 & GLAC1* IMPROVE site MONT1 IMPROVE site CABI1 & GLAC1* IMPROVE site SULA1 & SAWT1* IMPROVE site SULA1 IMPROVE site GAMO1 IMPROVE site MONT1 & GLAC1* IMPROVE site Lichens 12 reference sites – 2002 & 2003, 2010 3 reference sites – 1992, 2010 16 reference sites – 1992 - 1994, 2008 10 reference sites – 1992, 2008 3 reference sites – 2000 - 2001, 2008 4 reference sites – 2002, 2010 NADP Glacier NP MT05 NADP site* Glacier NP MT05* and Clancy MT07* NADP sites Glacier NP MT05* and Priest River ExpFst ID02* NADP sites Lost Trail Pass MT97 NADP site Clancy MT07* NADP site Glacier NP MT05 NADP site Snow Chemistry USGS snow chemistry sites* USFS R1 Class I Wilderness Areas AQRV’s

5. monitoring item ABWLMWGBWRWWCWGHW Lakes Twin Island & Stepping Stone Phase 3 lakes 3 lakes (2007)8 lakes (2007)5 lakes (2007) 3 lakes (2008) IMPROVE visibility YELL2* & NOAB1* IMPROVE sites YELL1* & NOAB1* IMPROVE sites GLAC1* IMPROVE site MONT1 & SULA1 IMPROVE sites SULA1 HECA1* SAWT1* IMPROVE sites Lichens 2 reference sites – 2008 NADP Yellowstone Pk WY08* NADP site Lost Trail Pass MT97 and Yellowstone Pk WY08* NADP sites Glacier NP MT95* NADP site Lost Trail Pass MT97 NADP site Lost Trail Pass MT97, Priest River Exp Fst ID02*, Palouse WA24* NADP sites Snow Chemistry USGS snow chemistry sites* USFS R1 Class II Wilderness Areas WAQV’s

7. Example of an IMPROVE baseline graph

8. R1 AIR QUALITY BEYOND EXCEL Excel limited in statistical powerExcel limited in statistical power Used SAS to run non parametric tests to test for statistically significant trends in USFS R1 AQ DataUsed SAS to run non parametric tests to test for statistically significant trends in USFS R1 AQ Data Analyzed R1 Lakes, NADP, and IMPROVE data.Analyzed R1 Lakes, NADP, and IMPROVE data.

9. Statistics SAS Institute statistical software was used to run analyses following draft USFS Data Analysis Protocol (DAP) recommendations in coordination with Lori Porth, RMRS Statistician Non-parametric test that can work with non-normal distributions and are not affected by errors, gross outliers, or missing data in the data set. A trend is detectable and considered significant if it meets our designated alpha level of α = 0.1 also shown as 90% confidence level. Additional confidence levels used were 95 (α = 0.05), 99 (α = 0.01), and 99.9 (α = 0.001). (Salmi et.al 2002).

10. Statistical Tests Used Mann-Kendall- run to see if there were significant trends for each parameter Kruskal-Wallace- run to see if seasons in the data set were statistically different Seasonal Mann-Kendall-run to look for trends while taking seasonality into account Sens slope estimator- magnitude of slope

11. USFS DAP Protocols

12. Our Hypothesis H o = Lake chemistry, air chemistry, and visibility show no trend through time H 1 = Lake chemistry, air chemistry, and visibility are not stable and have either an increasing or decreasing trend over time

13. R1 AQ Lake data limitations Mann-Kendall test was used for analysis of lakes. No seasonal data available More than 10 years of data available but from an array of months

14. Cabinet Mountain Wilderness Lower Libby Upper Libby Selway-Bitterroot Wilderness North Kootenai Shasta Absaroka-Beartooth Wilderness Stepping Stone Twin Island

16. Trends in Annual Lake pH

17. Trends in Annual Lake Conductivity

18. NADP sites in and around MT Glacier Lost Trail Pass Clancy Craters of the Moon Tower Junction Little Bighorn

20. Annual NH4+ at NADP sites

21. Annual Sulfate Concentration at NADP sites

22. Glacier NADP site percent confidence level and trend direction

23. Glacier NADP Spring Trends

24. Glacier NADP Seasonal Sulfate Trends

26. IMPROVE SITES GAMO1 SULA1 MONT1 GLAC1 CABI1 YELL2

27. Yell2 IMPROVE site on a clear day and on a hazy day Spectrum Series dv=0 Bext=10 SVR=390 Spectrum Series dv=17 Bext=52 SVR=75

28. IMPROVE Parameter Dictionary

29. Key IMPROVE Components PM2.5 components measured: –Sulfate (SO4) –Nitrate (NO3) –Organic Carbon (OMC) –Elemental Carbon (EC) also (LAC) –Coarse Particulate Matter (ECM) –Sea Salt –Fine Soils

31. Annual Trends in SVR at IMPROVE sites

32. YELL2 SVR best and worst days

33. YELL2 dv 20% best and worst days

34. Conclusions Trend interpretation, particularly cause/effect is difficult and complex Lake ANC decrease not statistically validated except at Stepping Stone Lake. The pH increasing trend and decrease in lake cation trends are not readily explainable Consistent NH4 increase trend at all of the NADP sites. This may be partially due to increased agriculture emissions such as feedlots in E. Oregon and E. Washington NO3 trend increases in lakes and NADP not as consistent as NH4 increase Consistent decrease SO4 at NADP sites is consistent with US trends the last 2 decades with reduced industrial sulfate emissions Consistent improvement in visibility at most of the IMPROVE sites as expressed in increased SVR, decreased deciviews, and reduced extinction 20% best and worst visibility day trends visually correlates well with wildfire emissions. More work in interpretation needs to be done before report finalization

35. Laurie Porth – RMRS Scott Copeland – USFS/CSU Lander Greg Bevenger – Shoshone NF Thomas Dzomba – USFS R1 Danke, Gracias, THANKS