Global Warming degrees over the last 50 years (e.g., IPCC)

Transcription

1

2

3 Global Warming degrees over the last 50 years (e.g., IPCC) Global Dimming (?) 10 20% decrease in surface solar radiation over the period Pan Evaporation Paradox If climate is warming, more energetic hydrologic cycle is expected implying an increase in evaporation. However, observations of pan evaporation across the U.S. and the globe show a decreasing trend in pan evaporation.

4 18 87 increasing increasing decreasing decreasing Warm-season pans Annual pans surface Trends in pan evaporation:

5

6

7

8 Mean annual Q N at 5-km resolution for WY

9 R G decreases over 97.5% of the U.S. Average decrease in R G of watts/m 2 /year, or 14.4% over 42 years.

10 Of the 44 annual pans, 64% show a decreasing ET pan trend. This is consistent with trends observed elsewhere [e.g., Chattopadhyay and Hulme, 1997; Peterson et al., 1995; Roderick and Farquhar, 2002] Of the 12 pans with significant trends (at the 90% level using the F-statistic of the trend slope), 75% are decreasing. Of the 228 warm-season pans, 60% show decreasing ET pan ; of the 43 pans with significant trends, 84% are decreasing. Warm-season ET pan has decreased across most of the conterminous U.S.

11 Traditional paradigm: ET pan independent of ET a ; ET a = f(et p, soil moisture)=min(et p, aθ c ) ET a Traditional Paradigm ETp Moisture Availability, θ

12 ET p is an increasing function of net available energy (i.e., of temperature) As the average global temperature increases, it is generally expected that the air will become drier and that evaporation will increase. ET a is an increasing function of ET p (according to traditional paradigm) ET p is decreasing globally while global temperature is increasing - a paradox Therefore, ET a should be decreasing

13 Two proposals for the decline in pan evaporation have been advanced: The first proposal is that pan evaporation has decreased because evaporation from the environment surrounding the pan has increased. However, if the proposed mechanism were the important one, then the vapor pressure deficit should have decreased. However, data from the United States show that its average has remained virtually constant over the past 50 years (10). The second invokes reductions in solar irradiance resulting from more clouds and/or aerosols and is generally consistent with the independent suggestion that increased pollution would weaken the hydrological cycle.

14 Bouchet's [1963] hypothesis states that over large homogeneous areas and away from sharp environmental discontinuities, there exists a complementary feedback mechanism between actual (ET a ) and potential (ET p ) evapotranspiration. ET w Wet Environment Evaporation - rate under conditions where the only limitation is the availability of energy; ET a Actual Evapotranspiration - occurs under conditions of limited moisture availability; ET p Potential Evapotranspiration - theoretical rate under conditions of limited moisture availability if the resulting excess in surface energy budget is used to evaporate further moisture. ET a + ET p = ket w

15

16 No theoretical proof of Bouchet s Hypothesis has been given Complementary Relationship models provide indirect evidence of its validity [e.g., Hobbins, Ram

17 Potential Evapotranspiration: Penman Equation: Drying Power of the Air, E A :

18 Wet Environment Evapotranspiration, ET w : Partial Equilibrium Evaporation Priestley and Taylor (1972) Actual Evapotranspiration, ET a :

19 Climatic data sets: Solar radiation - Solar and meteorological surface observation network (SAMSON) [NREL, 1993] Wind speed - SAMSON [NREL, 1993] Average temperature - Summary of the Day [NCDC] Dew-point temperature - Surface Airways [NCDC] Precipitation - Parameter-elevation Regressions on Independent Slopes Model (PRISM) [Daly et al., 1994] Summary of the Month [NCDC] and Summary of the Day [NCDC]: Annual: 44 pans, 1248 data. Warm-season: 228 pans, 7064 data. Geographic data sets: Elevation - DEM [NOAA-NGDC] Albedo - AVHRR [Gutman, 1988] Watershed boundaries - Hydro-climatic data network (HCDN) [Slack and Landwehr, 1992]

20 Pans Annual: 44 pans, 1248 data. Warm-season: 228 pans, 7064 data. Basins 655 basins 42 years

21 Thompsons 3 WSW, TX Death Valley, CA 300 (i) (ii) ETpan (mm) ETpan (mm) Months Months Homogeneous Non-homogeneous

22 Homogeneity test on 44 annual pans using nearest neighboring pans: Death Valley, CA Boulder City, NV ETpan (mm) ETpan (mm) Months Months

23 Homogeneity test on 44 annual pans using an a single a priori homogenous pan: Bartlett Dam, AZ Roosevelt, AZ ETpan (mm) ETpan (mm) Months Months

24 ETpan (mm) Months Before: trend = mm/year p = 7.35 * 10-12, R 2 = After: trend = mm/year p = 0.979, R 2 = ETpan (mm) Malheur Branch Exp Station, OR: 7 metadata-indicated heterogeneities: 5 station location shifts 2 of unknown origin Before After 200 Linear 0 (After) Linear Calendar year (Before)

25

26

27 Bouchet s Hypothesis implies that k equals 2. k is a function of both the stomatal resistance, r st, and the aerodynamic resistance, r a. k is strictly equal to 2 only for r st = 0 (i.e., for conditions of ample soil moisture) or for r a approaching infinity (i.e., k approaches 2 as the surface becomes smoother). Using ET pan as a surrogate for ET p, the value of k has a mean of 2.21 and a variance of Using ET p = k p ET pan, k has a mean of 1.97 and a variance of k p is the ratio of ET p /ET pan for the 192 estimates of ET pan (the average value of k p equals 0.83.) Such agreement of theory and observation is remarkable, especially for a data set covering such a large geographical area and long temporal span.

28 Observed Complementary Relationship Observed Complementary Relationship Annual Evapotranspiration (mm) ETpan ETw ETa* Annual Evapotranspiration (mm) ETp ETw ETa* Potential Humidity Index, F Potential Humidity Index, Φ

29 Observed Complementary Relationship Observed Complementary Relationship Annual Evapotranspiration (mm) ETpan ETp ETw ETa* Annual Evapotranspiration (mm) ETpan ETp Etw ETa* Potential Humidity Index, Φ Potential Humidity Index, Φ Pan ID Basin ID ID: Peace River at Arcadia, FL Observed Complementary Relationship Pan ID Basin ID : Salt River near Roosevelt, AZ Annual Evapotranspiration (mm) Potential Humidity Index, Φ ETpan ETp ETw ETa* Pan ID Basin ID : Guadalupe River near Spring Branch, TX

30 Szilagyi et al. [2001] Used a complementary relationship model to observe 6% increase in warm-season ET a Observe a 3% increase in budget-derived ET a Conclude that the complementary relationship overestimates trends in ET a Assume that dq n = 0, while cloudiness increases Also claim that dv dif = 0 Conclude that det pan = f(de a ) Problematic methodology gross averaging, use of ET pan for ET p

31 Roderick and Farquhar [2002] Examine two hypotheses to resolve pan evaporation paradox, described as ET pan decreasing while temperature rises: 1. Complementary relationship discounted as V dif remains constant over 50 years, so vapor transfer flux is unchanged 2. Decrease in irradiance Global Dimming observe a 10-20% decrease in Q n over 50 years. Conclude that det pan = f(dq n ) Problematic methodology spatial disjoint between ET pan and V dif observations.

32 ET w : ET p : ET w = f (Q n ) ET p ET pan = f (Q n, E A ) Trends in ET a: δet a = 2 ET w Q n ET p Q n δq ET p n E A δe A

33 ET p + Q n Q n = trends in local radiative flux E A = trends in regional advective flux ET rates ET w + E A E A ET a Q n Increasing moisture availability

34 Q n decreases over 97.5% of the conterminous U.S. Averaged decrease in Q n of watts/m 2 /year, or 14.4% over 42 years. Spatial mean decrease of 7.5% over 42 years Supports 50-year 10-20% decreases noted by Roderick and Farquhar, [2002]

35 Spatial mean decrease of 7.4 mm/yr/yr Trends are from either V dif or U 2, or both

36 Spatial mean decrease of mmhg over 42 years Refutes claims by Roderick and Farquhar [2002] and Szilagyi et al. [2001] that dv dif = 0

37

38 18 87 increasing increasing decreasing decreasing Warm-season pans Annual pans surface Trends in pan evaporation:

39

40 Trends in the radiative flux Q n (mm/yr/yr) ET pan > ET a < ET a > ET pan < 0 Trends in the vapor transfer flux E A (mm/yr/yr) -60

41

42 Bouchet s Hypothesis no longer a conjecture but an observational fact No paradox in the observations Simply an expression of complementarity between ET p (or ET pan ) and ET a Complementary Relationship explains >90% of all the basins observed (i.e., CR explains both increasing and decreasing ET p behavior.)

43 Examined distributed long-term, large-scale trends in ET a and in all of its components: observed distinct δq n decrease, supporting previous work, observed δe A not equal to zero, and δvdif not equal to zero, refuting previous work, dividing regional ET a trends up as to their origins shows that trends in BOTH the energy dynamic (Q n ) and the effects of regional advection (E A ) significantly affect trends in ET a, complementary relationship is an important dynamic in trends in ET a.

44 Specific: Three complicating factors in our homogenization procedure: data were gappy, metadata were incomplete, our definition of homogeneity was non-standard % of data at 75% of pans were adjusted for homogeneity. t-test assumes mean and variance of data in sub-periods can be described; counter to our assumption of trends in data. Conservative with errors. Conclusions of pan trend analyses not changed, changes in details of results, changes in respect to spatial analyses. Complementary relationship evident in adjusted data.