[ Pobierz całość w formacie PDF ]

Their main driving forces are the soil, firstly colluvium, secondly
limestone, and the east orientation..
24 GREAT BURGUNDY WINES  A PRINCIPAL COMPONENTS ANALYSIS
Figure 9  Clos-de-Vougeot Appelation, 2 First Components Map
4
S05
L2
2
L1
S13
L4
S08
S10
S09
S06
S04
S14
S01
-2
L3S02
-4
-4 -2 0 2 4
Component 1
Figure 10  Corton Appelation, 2 First Components Map
4
S05
L2
2
L1
S13
L4
S08
S10
S09
S06
S04
S14
S01
-2
L3S02
-4
-4 -2 0 2 4
Component 1
Component 2
�2004 by Frank Wittendal
Component 2
�2004 by Frank Wittendal
GREAT BURGUNDY WINES  A PRINCIPAL COMPONENTS ANALYSIS 25
Figure 11  Corton-Charlemagne Appelation, 2 First Components Map
4
S05
L2
2
L1
S13
L4
S08
S10
S09
S06
S04
S14
S01
-2
L3S02
-4
-4 -2 0 2 4
Component 1
Figure 12  Chevalier-Montrachet Appelation, 2 First Components Map
4
S05
L2
2
L1
S13
L4
S08
S10
S09
S06
S04
S14
S01
-2
L3S02
-4
-4 -2 0 2 4
Component 1
Component 2
�2004 by Frank Wittendal
Component 2
26 GREAT BURGUNDY WINES  A PRINCIPAL COMPONENTS ANALYSIS
Figure 13  Burgundy White Wines: Grands Crus, 2 First Components Map
4
S05
L2
2
L1
S13
L4
S08
S10
S09
S06
S04
S14
S01
-2
L3S02
-4
-4 -2 0 2 4
Component 1
Figure 14  Burgundy Red Wines: Grands Crus, 2 First Components Map
4
S05
L2
2
L1
S13
L4
S08
S10
S09
S06
S04
S14
S01
-2
L3S02
-4
-4 -2 0 2 4
Component 1
Component 2
�2004 by Frank Wittendal
Component 2
�2004 by Frank Wittendal
GREAT BURGUNDY WINES  A PRINCIPAL COMPONENTS ANALYSIS 27
Technical information on the analysis
able 4 (Principal components Weights) shows the equations of
T the principal components. For example, the first principal
component has the equation:
-0.053195*S14 - 0.388905*S13 - 0.0330027*S10 + 0.0648303*S09
-0.175634*S08 - 0.0647756*S06 + 0.406344*S05 - 0.0178382*S04
-0.0522355*S02 - 0.0865762*L1 - 0.506881*L2 - 0.123553*L3
- 0.59 4222*L4+0.0680913*S01
where the values of the variables in the equation are standardised
by subtracting their means and dividing by their standard devia-
tions.
Table 4  Component Weights
Component Component Component Component Component Component Component
1234567
------------ ------------ ------------ ------------ ------------ ------------ ------------
S14 -0.053195 -0.252398 -0.433137 0.201634 -0.251239 -0.262687 -0.432129
S13 -0.388905 0.139424 0.281544 -0.306742 -0.31667 -0.338916 0.0153694
S10 -0.0330027 -0.0774066 -0.336835 -0.334307 -0.236043 0.201487 -0.0739724
S09 0.0648303 -0.101651 0.0501833 -0.294474 -0.271963 0.0815049 0.0905124
S08 -0.175634 0.0316343 -0.319424 -0.0734412 0.278062 0.434561 0.452886
S06 -0.0647756 -0.177621 0.274006 0.165744 -0.223335 0.725516 -0.316859
S05 0.406344 0.480928 -0.0104936 0.373019 0.134512 -0.0649509 -0.0057148
S04 -0.0178382 -0.207238 -0.0567712 0.145756 -0.103976 -0.0665354 0.521234
S02 -0.0522355 -0.430396 0.107814 -0.270972 0.662858 -0.147659 -0.170871
L1 -0.0865762 0.214409 -0.618645 -0.155725 -0.0331862 0.0036301 -0.0258599
L2 -0.506881 0.342564 0.0898113 0.229818 0.008161 0.0171902 -0.0274139
L3 -0.123553 -0.40677 -0.150541 0.525338 -0.0053167 -0.0759619 0.0104742
L4 -0.594222 0.0365371 -0.0146438 0.178801 0.157673 -0.008311 0.0138379
S01 0.0680913 -0.279971 0.106848 0.122808 -0.292197 -0.13068 0.43729
Component Component
89
------------ ------------
S14 0.297558 -0.140284
S13 0.0206586 0.0044483
S10 -0.556056 0.329009
S09 0.422626 -0.41736
S08 0.397483 0.047094
S06 -0.0383282 -0.0520269
S05 -0.0371649 0.0265841
S04 -0.462846 -0.568305
S02 -0.0512563 0.0266962
L1 -0.0055177 -0.0082987
L2 -0.0271513 0.0236042
L3 -0.0517569 0.0497869
L4 -0.0254366 0.0104671
S01 0.203546 0.604429
28 GREAT BURGUNDY WINES  A PRINCIPAL COMPONENTS ANALYSIS
Table 5  Summary of Analysis
Percent of Cumulativ e
Number Eigenv alue Variance Percentage
1 2.34993 16.785 16.785
2 1.59703 11.407 28.193
3 1.33632 9.545 37.738
4 1.17063 8.362 46.099
5 1.09365 7.812 53.911
6 1.05883 7.563 61.474
7 1.03057 7.361 68.835
8 1.02272 7.305 76.141
9 1.01605 7.257 83.398
10 0.943622 6.74 90.138
11 0.699392 4.996 95.134
12 0.565501 4.039 99.173
13 0.0840012 0.6 99.773
14 0.0317635 0.227 100
he purpose of the principal components analysis is to obtain a
Tsmall number of linear combinations of the 14 variables which
account for most of the variability in the data. In this case, 9 com-
ponents have been extracted, since 9 components had eigenvalues
greater than or equal to 1.0. Together they account for 83.398% of
the variability in the original data.
he  scree plot (figure 15) shows that from a practical view point
Tonly the 3 first components are usefull for the purpose of our
investigation.
Figure 15  Contribution of components (% of data variance)
20
Scree Plot
15
10
5
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Component
Percent of variance
GREAT BURGUNDY WINES  A PRINCIPAL COMPONENTS ANALYSIS 29
Table 6  Alphabetical List of Burgundy  Grands Crus
B�tard-Montrachet
Bienvenues-B�tard-Montrachet
Bonnes-Mares
* Chablis Grand Cru
Chambertin
Chambertin-Clos de B�ze
Chapelle-Chambertin
Charlemagne
Charmes-Chambertin
Chevalier-Montrachet
Clos de la Roche
Clos de Tart
Clos de Vougeot
Clos des Lambrays
Clos Saint Denis
Corton
Corton-Charlemagne
Criots-B�tard-Montrachet
�chezeaux
Grands �chezeaux
Griotte-Chambertin
La Grande Rue
La Roman�e
La T�che
Latrici�res-Chambertin
Mazis-Chambertin
Mazoy�res-Chambertin
Montrachet
Musigny
Richebourg
Roman�e-Conti
Roman�e-Saint-Vivant
Ruchottes-Chambertin
* Not in study (vineyards of
Chablisien).
30 GREAT BURGUNDY WINES  A PRINCIPAL COMPONENTS ANALYSIS
References
Ashenfelter O., Storchmann K.  The quality of vineyard sites in
the Mosel valley of Germany  Proceedings of the R
nometrics X
conference, Budapest, May 23, 2003.
BRGM  Carte g�ologique harmonis�e  BRGM Orl�ans (France).
Carbonneau A.  Recherche sur les syst�mes de conduite de la
vigne : essai de ma�trise du microclimat et de la plante enti�re pour
produire �conomiquement du raisin de qualit�  Th�se de Docteur
Ing�nieur en R
nologie-Amp�logie, Universit� de Bordeaux II (France),
1980.
Haynes Simon J.  Concept of Terroir and the role of Geology 
In: Geoscience Canada, November 1999.
Lebon E.  De l influence des facteurs p�do- et m�soclimatiques sur
les comportements de la vigne et les caract�ristiques du raisin.
Application � l �tablissement de crit�res de zonage des potentialit�s
qualitatives en vignoble � climat semi-continental (Alsace)  Th�se
Universit� de Dijon (France), 1993.
Laville P.  Unit�s de terroir naturel et terroir. Une distinction
n�cessaire pour redonner plus de coh�rence au syst�me d appella-
tion d origine  In: Bulletin de l Office International de la Vigne et
du Vin n�745-746 mars-avril 1993  Paris (France) 2003.
Morlat R.  Terroirs viticoles : Etude et valorisation  R
noplurimedia
2001  Chaintr� (France) 2001.
Meinert L. D.  Terroir of the Red Mountain Appellation, Central
Washington State .  In: Geoscience Canada, December 2002, vol-
ume 29, number 4.
Pommerol C. The Wines and Wineland of France  Geological
journeys  Editions du BRGM (Orl�ans)  English translation
Robertson McCarta Ltd  London (UK), 1989.
Vaudour E.  Les Terroirs viticoles  D�finitions, caract�risation et
protection Dunod  Paris (France), 2003.
Wilson J. E.  The Origin and Odyssey of Terroir  In Geoscience
Canada Volume 26 Number 3, September 2001.
GREAT BURGUNDY WINES  A PRINCIPAL COMPONENTS ANALYSIS 31
Wilson J.E.  Terroir: the Role of Geology, Climate and Culture in
the making of French Wines  Mitchell Beasley London (UK)
1998.
Multivariate methods
PLS methodology, at http://ull.chemistry.uakron.edu/chemomet-
rics/multivariate_calibration/
Principal components methodology, at http://www.statsoftinc.com/text-
book/stfacan.html
Tennenhaus M.  La r�gression PLS g�n�ralis�e  In Cahiers de
recherche du groupe HEC  Jouy en Josas (France), 1993.
Volle M.  Analyse des Donn�es  Economica  Paris (France)
1993.
Wold. S. Geladi, P, Esbensen, K and �hman, J.  Multiway Principal
Components and PLS-Analysis  In Journal of Chemometrics 1,
41-56 (1987). [ Pobierz całość w formacie PDF ]