Highly regioselective amination of unactivated alkanes by hypervalent sulfonylimino-δ3-bromane

Article abstract of DOI:10.1126/science.1201686

Amination of alkanes has generally required metal catalysts and/or high temperatures. Here we report that simple exposure of a broad range of alkanes to N-triflylimino-δ³-bromane 1 at ambient temperature results in C-H insertion of the nitrogen functionality to afford triflyl-substituted amines in moderate to high yields. Marked selectivity for tertiary over secondary C-H bonds was observed; primary (methyl) C-H bonds were inert. Addition of hexafluoroisopropanol to inhibit decomposition of 1 dramatically improved the C-H amination efficiencies. Second-order kinetics, activation parameters (negative activation entropy), deuterium isotope effects, and theoretical calculations suggest a concerted asynchronous bimolecular transition state for the metal-free C-H amination event.

Full text of DOI:10.1126/science.1201686

Highly Regioselective Amination of Unactivated Alkanes by
Hypervalent Sulfonylimino- λ3-Bromane
Masahito Ochiai et al.
Science 332, 448 (2011);
DOI: 10.1126/science.1201686
This copy is for your personal, non-commercial use only.
If you wish to distribute this article to others, you can order high-quality copies for your
colleagues, clients, or customers by clicking here.
Permission to republish or repurpose articles or portions of articles can be obtained by
following the guidelines here.
The following resources related to this article are available online at
www.sciencemag.org (this information is current as of September 13, 2012 ):
Updated information and services, including high-resolution figures, can be found in the online
version of this article at:
http://www.sciencemag.org/content/332/6028/448.full.html
Supporting Online Material can be found at:
http://www.sciencemag.org/content/suppl/2011/04/20/332.6028.448.DC1.html
This article cites 26 articles, 1 of which can be accessed free:
http://www.sciencemag.org/content/332/6028/448.full.html#ref-list-1
This article appears in the following subject collections:
Chemistry
http://www.sciencemag.org/cgi/collection/chemistry
Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the
American Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. Copyright
2011 by the American Association for the Advancement of Science; all rights reserved. The title Science is a
registered trademark of AAAS.

REPORTS
iodanes PhI=NSO2Ar (7–10), which requires ac-
tivation by a transition-metal catalyst (Rh, Cu, or
Ag), and with amination by p-toluenesulfonyl
azide, which requires heating to 165°C to gen-
erate the reactive nitrene (19).
Five-, seven-, and eight-membered cyclic alkanes
similarly afforded high yields of N-triflylamides
3 at room temperature under metal-free condi-
tions. Somewhat decreased yields were obtained
in the reaction of cyclodecane (2e) and cyclo-
pentadecane (2f) (Fig. 1B).
Highly Regioselective Amination of
Unactivated Alkanes by Hypervalent
Sulfonylimino-l3-Bromane
Masahito Ochiai,1* Kazunori Miyamoto,1 Takao Kaneaki,1 Satoko Hayashi,2 Waro Nakanishi2*
Amination of alkanes has generally required metal catalysts and/or high temperatures. Here we
report that simple exposure of a broad range of alkanes to N-triflylimino-l3-bromane 1 at ambient
temperature results in C–H insertion of the nitrogen functionality to afford triflyl-substituted
amines in moderate to high yields. Marked selectivity for tertiary over secondary C–H bonds was
observed; primary (methyl) C–H bonds were inert. Addition of hexafluoroisopropanol to inhibit
We next explored the scope of aminations
using 1 with acyclic and substituted cyclic al-
kanes (Table 1). Acyclic alkanes 2 g-l afforded
moderate to good yields (34 to 74%) of amines 3,
decomposition of 1 dramatically improved the C–H amination efficiencies. Second-order kinetics, whereas 2,2,4,4-tetramethylpentane (2m) was re-
activation parameters (negative activation entropy), deuterium isotope effects, and theoretical
calculations suggest a concerted asynchronous bimolecular transition state for the metal-free
C–H amination event.
covered unchanged, though decomposition of
imino-l3-bromane 1 was observed in this case,
with formation of TfNH2 (91%) and p-CF3C6H4Br
(89%) (Table 1, entry 13). The use of excess
irect and selective functionalization of sulfonyl)imino-l3-bromane 1(15); the iminobromane amounts (20 equivalents relative to 1) of hexa-
alkane C–H bonds constitutes a long- 1 serves as an active organo nitrenoid species and fluoroisopropanol (HFIP) as an additive dramat-
standing goal in synthetic organic chem- directly undergoes stereospecific aziridination ically improved the C–H amination efficiencies,
D
istry (1–5). Alkanes are major constituents of of olefins with retention of stereochemistry and affording good to excellent yields of amides 3
petroleum and natural gas, but they are relatively transylidation to iodobenzenes, pyridines, and sul- (compare Table 1 entries 3/4, 5/6, 7/8, and 9/10)
inert and tend to be functionalized on a com- fides (16). In marked contrast to the reactions of while maintaining similar site selectivities; even
mercial scale through an inefficient sequence of imino-l3-iodanes, all of these reactions do not 2,3-dimethylbutane (2l) with a sterically encum-
overoxidation followed by reduction. Over the require the use of any transition-metal catalysts bered 3° C–H site afforded a high yield of ami-
past several decades, however, substantial pro- and proceed smoothly, even at room temperature. nation product in the presence of HFIP (entry 12).
gress has been reported, primarily through the These results suggest that 1 might be a more ac- Computational and kinetic studies suggest that
use of transition metal–catalyzed processes.
We focus here on the selective transformation perhaps even obviating the need for a catalyst.
of ubiquitous but unactivated aliphatic (sp3) C–H N-(trifluoromethylsulfonyl)imino-l3-bromane plex through hydrogen-bonding interactions (see
tive aminating agent of aliphatic C–H bonds, HFIP slows down the rate of decomposition of
imino-l3-bromane 1 via formation of a 1:1 com-
bonds to amines and amides. Both free-nitrene 1 was prepared by stoichiometric ligand exchange figs. S1 to S3 and SOM text) (20).
insertions and metal-nitrenoid aminations have of the fluorines in aryl(difluoro)-l3-bromane
In n-hexane (2g), amination of the secondary
been reported (Fig. 1A); in general, however, free- p-CF3C6H4BrF2 with triflylamide (TfNH2) in ace- C2-H and C3-H bonds occurred with comparable
nitrene processes are unselective and uncontrol- tonitrile at 0°C in a polytetrafluoroethylene per- selectivities. In branched alkanes, however, a no-
lable and, as a result, are of little synthetic value. fluoroalkoxy (Teflon PFA, DuPont, Wilmington, table preference for tertiary over secondary C–H
Therefore, the discovery by Breslow and Gellman DE) vessel (15, 17, 18). Compound 1 is sparingly bonds was observed in most cases, except in the
of the capacity of Mn and Fe tetraphenylporphyrin soluble in cyclohexane (2b) but smoothly trans- reactions of norbornane (2o) and trans-decalin (2t).
complexes to catalyze the tosylamination of cy- fers the nitrene unit (NTf) to the alkane. Stirring a No primary (methyl) C–H reactivity was observed
clohexane (2b) using N-tosylimino-l3-iodane was 0.01 M suspension in cyclohexane at ambient tem- in any of the substrates examined. Differences in
a promising development (6). Recently, Fiori et al. perature gradually dissolves the bromane 1, C–H bond dissociation energies account reason-
and some other groups greatly extended this meth- leading to selective insertion of the nitrogen func- ably well for these site selectivities (fig. S4 and
odology and have found an efficient method for tionality into the unactivated methylene C–H SOM text). Norbornane (2o) exclusively under-
Rh-catalyzed high-yield amination of 2b using group. After stirring for 24 hours, N-(cyclohexyl) goes C–H insertion at the 2° C2-H site (with high
in situ–generated imino-l3-iodane, indicating the triflylamide (3b) was isolated in 91% yield (Fig. exo selectivity), though in this case the 3° C1-H
powerful nature of the transition metal–catalyzed 1B). This reactivity is in marked contrast with the bond is strengthened by the rigid structure around
amination of aliphatic C–H bonds (7–10).
reported amination of cyclohexane using imino-l3- C1. A similar preference for oxyfunctionalization
Hypervalent aryl-l3-bromanes exhibit higher
reactivity compared with aryl-l3-iodanes (the l3
notation indicates that the bromine and iodine
centers exceed their standard valence by two) (11).
This is probably due to the greater electroneg-
ativity as well as the larger ionization potential of
bromine relative to iodine (12–14). Recently, we
reported synthesis of hypervalent N-(trifluoromethyl-
Fig. 1. (A) Strategy for
RN
A
unactivated C–H amination.
R, arenesulfonyl, alkanesul-
fonyl, alkoxysulfonyl, alk-
oxycarbonyl, etc. (B) C–H
amination of cycloalkanes with
N-triflylimino-l3-bromane
1 under metal-free condi-
tions at room temperature.
Yields shown in parentheses
were determined by GC;
otherwise, isolated yields
are reported. Tf = CF3SO2.
-
nitrene
RN Y
C-H
amination
organo nitrenoid
(Y: leaving group)
(metal free)
M NR
metal nitrenoid
(M: metal)
B
1Graduate School of Pharmaceutical Sciences, University of
Tokushima, 1-78 Shomachi, Tokushima 770-8505, Japan. 2De-
partment of Material Science and Chemistry, Faculty of Systems
Engineering, Wakayama University, 930 Sakaedani, Wakayama
640-8510, Japan.
-
+
TfN Br
Br
3a: n = 1 yield: (92%)
NHTf
( )n
3b:
3c:
3d:
3e:
3f:
2
3
4
6
91%
70%
25 °C
Ar
+
+
( )n
81%
(65%)
(54%)
CF3
1 (0.01 M)
CF3
11
*To whom correspondence should be addressed. E-mail:
mochiai@ph.tokushima-u.ac.jp
2
3
448
22 APRIL 2011 VOL 332 SCIENCE www.sciencemag.org

REPORTS
S14
S14
S14
Br1
Br1
Br1
N13
N13
N13
H22
H22
C21
C21
H22
C21
TS-1
TS-2
TS-3
r
C21-H22 = 1.183 Å
r
r
C21-H22 = 1.218 Å
H22-N13 = 1.291 Å
r
r
C21-H22 = 1.197 Å
H22-N13 = 1.348 Å
rH22-N13 = 1.394 Å
rN13-Br1 = 2.173 Å
rN13-Br1 = 2.264 Å
rN13-Br1 = 2.203 Å
Fig. 2. Transition state structures for C–H aminations of ethane (TS-1), propane (TS-2), and isobutane (TS-3) with PhBr=NTf 4 optimized at the MP2/6-311+G
(d) (Br, S, N) and 6-31G(d,p) (F, O, C, H) levels. r, bond length.
-
+
Fig. 3. Deuterium kinetic
isotope effects for C–H ami-
nations. D, deuterium.
correction for differing numbers of H atoms in
each substrate, are shown in Table 1.
TfN Br
NHTf
NHTf
D11
25 0.2 °C
24 h, Ar
Adamantane (2v) is the most reactive sub-
strate that we examined. The magnitude of reac-
tivities of 3° C–H bonds relative to cyclohexane
(2b) varies widely, from an 81-fold enhancement
for adamantane (2v) to a 3.5-fold increase for
dimethylpentane (2j). These results again dem-
onstrate that, in molecules where C–H bonds of
similar electron densities are present, sterics can
play a major role in guiding selectivity.
The second-order kinetics for reaction of
1 with alkanes suggest that generation of free
nitrene is not rate limiting (and may not even be
on the pathway). Theoretical calculations on the
+
+
+
D12
CF3
kH kD = 3.72
/
2b
2b-d12
99.6% D
3b
yield 68%
3b-d11
yield 18%
1 (0.01 M)
D
kinetics:
2,2,4,4-tetramethylpentane (2m)/30 0.1 °C/Ar
D
D
D
kH kD = 1.49
/
2v-d4 94% D
2v
at C2 and for exo stereoselectivity was observed the relative rates of C–H insertion and associated reactions of phenyl(imino)-l3-bromane PhBr=NTf
in norbornane oxidation with dioxiranes (21) and regioselectivity.
oxaziridines (22). Silver-catalyzed amination of The stereochemistry of trans-3pa and cis- that the generation of singlet nitrene is higher in
simple alkanes with imino-l3-iodane showed a 3qa was determined by conversion of the Tf-N- energy than bimolecular C–H amination path-
4 with ethane, propane, and isobutane also predict
close parallel 3° > 2° > 1° selectivity (23).
substituted dimethylcyclohexanes to the known ways, in which iminobromane 4 functions as an
Steric hindrance due to the bulky aryl-l3- mesylamides through N-benzylation, b elimination active nitrenoid species (fig. S11). The calculated
bromanyl and triflyl groups attached to the of TfH, NaBH4 reduction, debenzylation, and fi- activation barriers (in kilocalories per mole) for
nitrogen atom of 1 also seems to influence the nally mesylation (figs. S6 and S7). These results C–H insertion with 4 showed a monotonic but
relative substrate reactivities and positional se- clearly indicated exclusive retention of the start- large decrease in the order Et-H (22.5), i-Pr-H
lectivities. For instance, the sterically congested ing stereochemistry of 2p and 2q in the products. (17.1), and t-Bu-H (13.1), which parallels the
methylene group of tetramethylated pentane 2m
Rates for C–H amination of alkanes with experimentally obtained relative reactivities of
undergoes no amination (facilitating the use of imino-l3-bromane 1 (9.5 × 10−5 M, where a clear C–H bonds. The TfN insertion (TS-2) into the
2m as an inert solvent for amination of solid solution of 1 was obtained) at 30°C were mea- secondary center in propane by 4 proceeds with
substrates such as 2o and 2v). Differences in sured spectrophotometrically under pseudo–first- a barrier of only 17.1 kcal mol−1, which is very
3°/2° site selectivities for trans- (2r) and cis-1,2- order conditions using inert tetramethylpentane close to the evaluated activation enthalpy (DH≠)
dimethylcyclohexane (2s) as well as trans- (2t) 2m as a solvent (tables S1 and S2). In all cases of 19.4 kcal mol−1 for the reaction of cyclohexane
and cis-decalin (2u) (with both cis-isomers examined, the observed rate constants kobs proved (2b) with imino-l3-bromane 1 (fig. S8). A moder-
showing higher 3° selectivity) probably reflect proportional to the concentration of alkane and ate and negative activation entropy (DS≠= –8.38
differences in steric hindrance and/or strain re- the slopes of such plots afforded the second-order cal mol−1K−1) for the amination of cyclohexane
lease on approaching the transition state of the rate constants (k2/M−1s−1): 0.000778 (2a), 0.000989 (2b) indicates some loss of mobility in the transi-
amination (Table 1, entries 18 to 21) (24). These (2b), 0.00127 (2o), and 0.0289 (2v). The ob- tion state, in good agreement with a concerted bond-
considerations were supported by competition served second-order rate law suggests the likely formation/scission process obtained by calculations.
experiments that exposed 1 simultaneously to involvement of a bimolecular transition state. For
In each of the calculated C–H amination tran-
trans- (2p) and cis-1,4-dimethylcyclohexane each compound, second-order rate constants k2 sition states TS-1 (ethane insertion), TS-2 (pro-
(2q): The equatorial tertiary C–H bond reacts (CH2) for methylene groups and k2 (CH) for meth- pane CH2 insertion), and TS-3 (isobutane CH
4.4 times more rapidly than those oriented ax- ine groups were calculated based on the product insertion), the alkane moiety (RH, where R is
ially (fig. S5). Hence, a subtle balance of elec- profiles shown in Table 1; relative reactivities Et, i-Pr, t-Bu) becomes positively charged (by
tronic and steric factors seems to determine of each hydrogen atom, obtained after statistical +0.2806, +0.2235, and +0.1914, respectively),
www.sciencemag.org SCIENCE VOL 332 22 APRIL 2011
449

REPORTS
whereas the total positive charge (+0.8816) of suggest transfer of alkane C–H s-bonding elec- activation and the N13-C21 bond formation take
the PhBr moiety in 4decreases to +0.3754, +0.4524, trons to iminobromane 4 with some hydride- place as a single event, TS-1 to TS-3 indicate
and +0.4896, respectively (Fig. 2). These results transfer character. Although the C21-H22 bond a highly asynchronous transition state, with
N13-H22 bond formation being much more ad-
vanced than N13-C21 bond formation (Fig. 2).
Table 1. Metal-free C–H amination at room temperature with imino-l3-bromane 1. Unless otherwise noted,
These transition-state structures are reminiscent
0.01 M of 1 (with 20 equivalents of HFIP, where indicated) was stirred in neat alkane under Ar for 10 to 24
of those calculated for the concerted alkane C–H
hours. Yields and ratios of the indicated products 3 (major isomer on the left) were determined by GC
(numbers in parentheses are isolated yields of the major product; N = NHTf). Relative reactivities of C–H
bonds compared with cyclohexane (2b) after statistical correction (per H atom) are shown in red (methine
insertions by dimethyldioxirane (25) and rhodium-
carbene/nitrene complex (26, 27). As observed
for trans- (2p) and cis-1,4-dimethylcyclohexane
(2q), the aminations proceed with full retention
of stereochemistry, indicating that long-lived car-
bocation or free radical intermediates are not in-
volved in the aminations.
groups) and blue (methylene groups).
Deuterium kinetic isotope effects (KIEs) were
evaluated by the competitive reaction of imino-
l3-bromane 1 with a 1:1 mixture of cyclohexane
(2b) and cyclohexane-d12 (2b-d12) at 25°C: A
primary deuterium KIE kH/kD (kH, rate constant
of 2b; kD, rate constant of 2b-d12) of 3.72 for 2°
C–H insertion was determined by gas chroma-
tography (GC) (Fig. 3). Comparison of k2 values
(table S1) for 3° C–H insertion of adamantane
(2v) and 1,3,5,7-tetradeuterioadamantane (2v-d4)
in tetramethylpentane 2m at 30°C afforded a
smaller kH/kD value of 1.49 (28). These moderate
KIE values suggest only partial C–H bond break-
ing in the transition state and are not compatible
with a hydrogen atom abstraction/radical-rebound
mechanism, as proposed for Ru- and Cu-nitrenoid–
based C–H aminations with large KIEs of 6.1 to
6.6 (29, 30). The smaller KIE value observed for
the tertiary C–H bonds of 2v relative to the sec-
ondary C–H bonds of 2b is consistent with the
calculations, which show greater extension of the
breaking C21-H22 and N13-Br1 bonds in TS-2
than in TS-3 (Fig. 2); the latter structure reflects
the more early character of the transition state.
We feel that these KIEs—together with the second-
order kinetics, kinetic parameters (negative acti-
vation entropy), retention of stereochemistry, and
theoretical calculations—strongly implicate a con-
certed asynchronous pathway for metal-free C–H
amination with sulfonylimino-l3-bromane 1. The
driving force for this room-temperature uncata-
lyzed insertion into unactivated alkane C–H
bonds is probably the greatly enhanced nucleofu-
gality of aryl-l3-bromanyl groups compared with
aryl-l3-iodanyl groups, which correlates with
increased instability of hypervalency at bromine
relative to iodine (31–33).
References and Notes
1. J. A. Labinger, J. E. Bercaw, Nature 417, 507 (2002).
2. K. Godula, D. Sames, Science 312, 67 (2006).
3. A. A. Fokin et al., J. Am. Chem. Soc. 122, 7317 (2000).
4. H. M. L. Davies, J. R. Manning, Nature 451, 417 (2008).
5. P. R. Schreiner, A. A. Fokin, Chem. Rec. 3, 247 (2004).
6. R. Breslow, S. H. Gellman, J. Chem. Soc. Chem. Commun.
1982, 1400 (1982).
7. M. R. Fructos, S. Trofimenko, M. M. Díaz-Requejo,
P. J. Pérez, J. Am. Chem. Soc. 128, 11784 (2006).
8. Z. Li, D. A. Capretto, R. Rahaman, C. He, Angew. Chem.
Int. Ed. 46, 5184 (2007).
9. K. W. Fiori, J. Du Bois, J. Am. Chem. Soc. 129, 562 (2007).
10. C. Liang et al., J. Am. Chem. Soc. 130, 343 (2008).
11. U. Farooq, A.-H. A. Shah, T. Wirth, Angew. Chem. Int. Ed.
48, 1018 (2009).
450
22 APRIL 2011 VOL 332 SCIENCE www.sciencemag.org

REPORTS
12. M. Ochiai, Synlett 2009, 159 (2009).
13. Ionization potentials of halobenzenes increase in
the order PhI (8.69 eV) < PhBr (8.98 eV) < PhCl
(9.06 eV) (14).
14. D. R. Lide, Ed., CRC Handbook of Chemistry and Physics
(CRC Press, Boca Raton, FL, 1992).
15. M. Ochiai et al., J. Am. Chem. Soc. 129, 12938
(2007).
16. M. Ochiai, K. Miyamoto, S. Hayashi, W. Nakanishi,
Chem. Commun. 46, 511 (2010).
22. D. D. DesMarteau, A. Donadelli, V. Montanari,
V. A. Petrov, G. Resnati, J. Am. Chem. Soc. 115, 4897 (1993).
23. B. P. Gómez-Emeterio, J. Urbano, M. M. Diaz-Requejo,
P. J. Perez, Organometallics 27, 4126 (2008).
24. K. Chen, A. Eschenmoser, P. S. Baran, Angew. Chem. Int.
Ed. 48, 9705 (2009).
31. M. Ochiai, N. Tada, T. Okada, A. Sota, K. Miyamoto,
J. Am. Chem. Soc. 130, 2118 (2008).
32. T. Okuyama, T. Takino, T. Sueda, M. Ochiai, J. Am.
Chem. Soc. 117, 3360 (1995).
33. P. J. Stang, Z. Rappoport, M. Hanack, L. R. Subramanian,
Vinyl Cations (Academic Press, New York, 1979).
Acknowledgments: This work was supported by a
Grant-in-Aid for Scientific Research (B) (Japan Society
for the Promotion of Science). We thank Central Glass
Co., Japan, for a generous gift of BrF3.
25. M. N. Glukhovtsev, C. Canepa, R. D. Bach, J. Am.
Chem. Soc. 120, 10528 (1998).
26. E. Nakamura, N. Yoshikai, M. Yamanaka, J. Am.
Chem. Soc. 124, 7181 (2002).
17. Teflon PFA is a perfluoroalkoxy copolymer resin, available
27. X. Lin, C. Zhao, C.-M. Che, Z. Ke, D. L. Phillips,
Supporting Online Material
www.sciencemag.org/cgi/content/full/332/6028/448/DC1
Materials and Methods
SOM Text
Figs. S1 to S12
Tables S1 and S2
from DuPont, Wilmington, Delaware.
18. Materials and methods are available as supporting
material on Science Online.
19. D. S. Breslow, M. F. Sloan, N. R. Newburg, W. B. Renfrow,
J. Am. Chem. Soc. 91, 2273 (1969).
20. C. Reichardt, Solvents and Solvent Effects in Organic
Chemistry (Wiley-VCH, Weinheim, Germany, 2003).
21. R. Mello, M. Fiorentino, C. Fusco, R. Curci, J. Am. Chem.
Soc. 111, 6749 (1989).
Chem. Asian J. 2, 1101 (2007).
28. Small deuterium KIEs of 1.05 and 1.07 for the hydride
abstraction reaction of the bridgehead adamantane C–H
bond with tert-butyl cation and hydroperoxonium ion
were reported (3).
29. S.-M. Au, J.-S. Huang, W.-Y. Yu, W.-H. Fung, C.-M. Che,
J. Am. Chem. Soc. 121, 9120 (1999).
References
30. Y. M. Badiei et al., Angew. Chem. Int. Ed. 47, 9961
(2008).
14 December 2010; accepted 4 March 2011
10.1126/science.1201686
period. Because the seasonal cycle typically is
large, care must be exercised in determining trend
information from the data set [supporting online
material (SOM)].
We aim here to determine whether there is a
statistically significant trend [where trend is de-
fined as a linear increase or decrease in the mean
(23, 24)] within the time series of monthly mean,
90th-, and 99th-percentile values of wind speed
and wave height for 2° × 2° regions covering the
globe. In the analysis, we take particular care to
ensure that the trend can be separated from the
seasonal component (SOM).
Global Trends in Wind Speed
and Wave Height
I. R. Young,* S. Zieger, A. V. Babanin
Studies of climate change typically consider measurements or predictions of temperature over
extended periods of time. Climate, however, is much more than temperature. Over the oceans,
changes in wind speed and the surface gravity waves generated by such winds play an
important role. We used a 23-year database of calibrated and validated satellite altimeter
measurements to investigate global changes in oceanic wind speed and wave height over this
period. We find a general global trend of increasing values of wind speed and, to a lesser
degree, wave height, over this period. The rate of increase is greater for extreme events as
compared to the mean condition.
The trend was quantified as a linear function
over the duration of the time series. The analysis
revealed that 90th- and 99th-percentile wind speed
data from the GEOSAT altimeter were of ques-
ceanic wind speed and wave height help speed and wave height, using a variety of instru- tionable quality (SOM). Therefore, these data were
to control the flux of energy from the ments, including altimeters, scatterometers, and excluded from the analysis. As a result, the wave
atmosphere to the ocean (1) and upper synthetic aperture radar, providing global cover- height analysis considers the period 1985–2008,
O
ocean mixing (2). Thus, they substantially influ- age of wind and/or waves. Of these instruments, whereas wind speed is analyzed for the shorter
ence the mechanisms of air-sea interaction (3). the radar altimeter provides by far the longest- period 1991–2008. The trend was expressed for
Previous attempts to investigate trends in oceanic duration record. Since the launch of GEOSAT in each 2° × 2° region as the annual percentage in-
wind speed and wave height have used ship ob- 1985, there exists an almost continuous (there crease or decrease relative to the mean condi-
servations (4–8), point measurements (9), numer- was a break in 1990–1991) record of measures tion and in absolute terms. The monthly mean,
ical modeling (10–15), or satellite observations from a total of seven different altimeter missions. 90th-, and 99th-percentile trend values for both
(16). Almost all of these studies are regional Numerous calibrations of these altimeters have wind speed and wave height are shown in Figs.
rather than global. Although there is a range of shown that the instruments can be used to mea- 1 to 3 (percentage increase or decrease) and figs.
pffiffiffi
results, many studies show an increasing trend in sure significant wave height, Hs ¼ 4 E, where S7 to S9 (SOM) (absolute increase or decrease),
significant wave height, particularly in the North E is the total energy of the wave field, with a root respectively.
Atlantic and North Pacific, often correlated with mean square (rms) error of less than 0.2 m (17),
There is a clear global increase in wind
interannual variations such as the North Atlantic and wind speed, U10, with a rms error of less than speed for all three statistics. The mean and 90th-
Oscillation. Careful ship observations (4–6) also 1.5 m/s (17–21). Data from altimeter missions percentile wind speed trends are relatively sim-
show waves locally generated by the wind (hereafter have been used to investigate mean ocean wind ilar, with the magnitude of the increase being
referred to as wind-sea) and swell behaving quite and wave climatology (21, 22) on a global scale. larger for the 99th percentile. Such a result in-
differently and that there exist quite different trends Recently, Zieger et al. (20) carried out systematic dicates that the intensity of extreme events is in-
in wind speed and wave height. The present analysis calibrations and cross-platform validations of all creasing at a faster rate than that of the mean
uses recently developed satellite altimeter data sets altimeter measurements over the full 23 years for conditions. At the mean and 90th percentile, wind
to carefully investigate such trends on a global scale. which data are available. This study provided a speeds over the majority of the world’s oceans
Satellite-based systems provide an alternative consistent data set over this extended period. Be- have increased by at least 0.25 to 0.5% per year
to visual or in situ measurements of oceanic wind cause the data set spans multiple satellite plat- (a 5 to 10% net increase over the past 20 years).
forms, consistent calibration and validation are The trend is stronger in the Southern Hemisphere
critical when investigating long-term trends. In than in the Northern Hemisphere. The only sig-
Swinburne University of Technology, Melbourne, Victoria,
the present study, we used this data set to inves- nificant exception to this positive trend is the cen-
tigate whether there have been systematic changes tral north Pacific, where there are smaller localized
in the ocean wind and wave climate over this increases in wind speed of approximately 0.25%
Australia.
*To whom correspondence should be addressed. E-mail:
ir.young@anu.edu.au
www.sciencemag.org SCIENCE VOL 332 22 APRIL 2011
451