Journal of Cultural Heritage 13 (2012) 365–370

Available online at

www.sciencedirect.com

Original article

Future climate-induced pressures on painted wood

Łukasz Bratasz a,∗ , Ian Harris b , Łukasz Lasyk a , Michał Łukomski a , Roman Kozłowski a
a
Institute of Catalysis and Surface Chemistry, Polish Academy of Sciences, ul. Niezapominajek 8, 30–239 Kraków, Poland
b
School of Environmental Sciences, University of East Anglia, Norwich NR4 7TJ, United Kingdom

a r t i c l e i n f o a b s t r a c t

Article history: A broad category of cultural heritage objects are multilayer structures composed of organic, humidity-
Received 20 April 2011 sensitive materials – wood, animal glue, paper, leather, bone or paints. They respond to variations in
Accepted 23 January 2012 relative humidity (RH) in their environment by cyclically gaining and losing moisture, and consequently
Available online 3 March 2012
swelling and shrinking. Differences in the moisture response of the materials induce internal stresses
in the individual layers of the structures, which cause objects to deform and crack. Polychrome wood
Keywords: is examined in detail. The cumulative physical damage of the design layer on wood due to repeated RH
Climate change
variations is quantified in terms of their magnitude and number of times they occur. The climatological
Relative humidity
Painted wood
risk index for accumulated, ‘fatigue’ damage is established, using a procedure to reduce irregular real-
Damage world climate histories into simple RH cycles of known damage impact. Using output from the Hadley
Risk index Model (HadCM3) and simple transfer functions predicting indoor temperature and RH from outdoor
climate, changes in the indoor climate through to 2100 were forecast for unheated buildings. European
maps highlighting the areas in which painted wood may be significantly affected by climate change are
presented.
© 2012 Elsevier Masson SAS. All rights reserved.

1. Research aims humidity-sensitive objects due to climate becoming more extreme

and unstable [3], which will require a broader use of environ-
The main objective of this study was to derive a risk index for the mental control systems or updating and expanding the capacity
climate-induced damage to the design layer on wood subjected to of existing ones. On the other hand, there has been a growing
complex climatic variations, by decomposing them mathematically movement to greening of museums, that is, finding low-carbon,
into simple relative humidity (RH) cycles of various sizes whose cost-effective solutions to the use of energy in the museums while
damaging impact was assessed in the laboratory. Using this risk at the same time maintaining high standards of the collection care
index and the changes in outdoor climate through to 2100 available [2]. There has been a universal agreement that the development
from the Hadley Model, future climate-induced pressures on the of environmental guidelines for cultural heritage needs to be
painted wood could be predicted. informed by growing scientific understanding of how changes in
environmental conditions ultimately affect real artefacts.
2. Introduction This paper focuses on the response of paintings on wooden
panels – an important category of the multilayer humidity-
A broad category of cultural heritage objects are multilayer sensitive objects – to variations in the environment. Materials that
structures composed of organic, humidity – sensitive materials – constitute painted wooden objects – wood substrate, animal glue,
wood, animal glue, paper, leather, bone or paints. Environmental gesso and paints – respond to the environmental variations by gain-
standards for their display or storage have been much debated ing moisture and swelling when RH is high and losing moisture
in the recent years. The transcriptions of two ‘round tables’ of and shrinking when the surrounding air is dry. The mismatch in
the International Institute for Conservation: ‘Climate Change and the dimensional response of gesso and wood, in its most respon-
Museum Collections’ in 2008 and ‘The Plus/Minus Dilemma: The sive tangential direction, has been identified as the worst case
Way Forward in Environmental Guidelines’ in 2010 illustrate condition: upon desiccation, the wood shrinkage overrides that of
the problems discussed and the recommendations on the way the gesso which experiences compression, whereas, upon wood
forward [1,2]. On one hand, the interest in the issue has been swelling, the gesso layer experiences tension [4]. If the elongation
driven by a predicted future increase in the risk of damage to the or contraction in the gesso go beyond critical levels, the material
can crack or delaminate, respectively.
Damage of painted wood due to the RH changes is not limited
∗ Corresponding author. Tel.: +48 12 63 95 11 9; fax: +48 12 42 51 92 3. to the pictorial layer only. The wood substrate may also experience
E-mail address: [email protected] (Ł. Bratasz). stress when prevented from freely swelling or shrinking across

1296-2074/$ – see front matter © 2012 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.culher.2012.01.013
366 Ł. Bratasz et al. / Journal of Cultural Heritage 13 (2012) 365–370

its grain. Restraint may result from rigid construction restricting A2(a) emission scenario which represents a very heterogeneous
wood’s movement like framing of a painting or an ill-designed rigid world characterised by a strengthening regional cultural identities,
structure (cradle) added to confer greater planarity and dimen- with an emphasis on family values and local traditions, with high
sional stability to a painting. Wood can also experience internal population growth [12]. The scenario gives pronounced changes
restraint as the moisture diffusion is not instantaneous and the in future climate. Detailed justification for using this model to
outer parts of the wood respond dimensionally more quickly than assess the overall risk posed to monumental heritage by climate
the interior to variations in ambient RH. Analysis of the described change, and information on the methodology of mapping predicted
stress sources in restrained wood substrates is a separate issue, damage for Europe is contained in Ref. [3], an outcome of a larger
which is outside the scope of this paper. European project, NOAH’s ARK. 30-day averages of temperature
and RH were used to estimate the indoor climate for three 30-year
3. Materials and methods periods: 1961–1990 (baseline), 2010–2039 (near future) and
2070–2099 (far future).
To produce specimens imitating historic panel paintings, lime
wood substrates 10 mm thick were sized with rabbit-skin glue and 4. Results and discussion
coated with gesso composed of the same glue and ground chalk
(the pigment) at pigment-volume concentration (PVC) of 92% com- 4.1. Structural response of painted wood to variations in ambient
monly used in the restoration of panel paintings: the thickness of relative humidity
the dried gesso layer was approximately 0.5 mm.
Cycles of elongation in the gesso were produced by mechanically 4.1.1. Critical elongations causing damage of the design layer
stretching and compressing the specimens, at selected ampli- The DSPI technique can provide quantitative information about
tudes, which imitated dimensional changes induced by repetitive the number of cracks in the design layer on wood and their length
fluctuations of RH. The frequency of the elongation cycles was [5,6]. Consequently, an assessment of the cumulative damage to
approximately 0.3 Hz, and the specimens were subjected up to the layer as a function of the number of gesso elongation cycles was
36,500 cycles, equivalent to approximately 100 years of diurnal possible. Plots of the cumulative crack length versus the number
climatic fluctuations producing the same elongation. An exten- of stretching cycles were determined for a range of elongations
someter continuously monitored the elongation in the specimens. between 0.0015 and 0.005. For elongation of 0.0015, no fracture
The experiments were conducted under a constant RH of 50% in gesso appeared after the maximum number of 36,500 cycles
maintained in the laboratory. The specimens were taken out after applied, whereas elongation of 0.0025 produced first cracking after
a predetermined number of elongation cycles and the fracture 5000 cycles only (Fig. 1). The results obtained allow plotting an
development in the gesso was recorded using digital speckle pat- S-N curve where S is the elongation (strain) leading to fracture and
tern interferometry (DSPI), which is capable of monitoring physical N is the number of cycles to cause the first incidence of fracture
damage at the micro-level before it is discernible visually. Further at that elongation (Fig. 2). The general curve shape is sigmoid,
details of the experimental approach used can be found in refer- starting from the elongation for fracture in a single cycle or a few
ences [5,6]. cycles, and dropping to a plateau where cyclic elongation can be
The critical elongation values were translated into the critical tolerated for up to 36,500 cycles, equivalent to 100 years of diurnal
magnitudes of RH variations necessary to cause these elongations elongation cycles. The elongation of 0.002 was assumed to be
by using the swelling/shrinkage responses of wood [7] and gesso close to that value. Thus, the elongation tolerable at the maximum
[8] determined experimentally. number of cycles was approximately 1/3 of the single cycle fracture
Climate predictions for the future have been extracted from the elongation.
output of two widely used models from the Hadley Centre, UK: It should be noted at this point, that the levels of elongation
HadCM3 and HadRM3. HadCM3 is a coupled ocean-atmosphere at which the first incidence of fracturing on undamaged gesso
global circulation model with a grid resolution of 2.5 × 3.75◦ i.e. appears, were accepted as the critical ones. This accepted criterion
278 × 295 km2 at 45◦ N latitude [9–11]. HadRM3 is a regional of damage is very conservative, as design layers on wooden historic
climate model, which encompasses Europe at a higher resolution
(a grid of equal-area cells, 50 × 50 km2 ) but spans only the years
2070 to 2099. The model outputs used in this work relied on the

Fig. 2. Elongation, and corresponding amplitude of relative humidity (RH) varia-

tion, leading to fracture in gesso versus number of cycles to cause fracture at that
elongation. The most responsive tangential direction in wood was considered and
each RH variation was assumed to cause a full response of an unrestrained panel.
Fig. 1. The cumulative crack length in gesso versus the number of elongation cycles. Amplitude of RH variation was calculated assuming a starting point of 50% RH.
 Ł. Bratasz et al. / Journal of Cultural Heritage 13 (2012) 365–370 367

in the gesso and small relative amount of glue dramatically reduce

the moisture adsorption, and the related dimensional response.
The critical magnitudes of RH variations necessary to cause the
critical elongations in gesso were calculated from the respective
moisture related elongations in the wood substrate and the gesso
determined as described above and are shown on the right axis
of the S-N curve in Fig. 3. The approach adopted is valid only for
the RH variations which last much longer than the response time
of a wooden panel so that the panel can reach the new values of
the equilibrium moisture content, and the related elongation, at
each instant of a variation. The response time of a panel depends
on its thickness and the configuration of water vapour diffusion,
that is, if the moisture flow is possible through both the faces
of a panel or just one, rear face, when the painted face is cov-
ered by an impermeable layer of paint or varnish. Typically, the
original thickness of panels ranged between 30 mm and 45 mm
[16]. However, flattening of deformed panels in the nineteenth
century frequently involved planning them to a fraction of their
Fig. 3. Average swelling and shrinkage isotherms at 24 ◦ C based on the data mea-
sured in the tangential direction for 21 wood species important for cultural heritage. original thicknesses. Therefore, panel paintings, even large-sized,
The intermediate isotherm was generated for the readsorption path from 31.5 to surviving to our times can be a mere 2–5 mm thick. However,
96% relative humidity (RH) and used to derive dimensional changes induced by RH such extremely thin panels were always backed by new sup-
variations around the mid-value of 50% RH (solid circles).
ports usually restraining their dimensional response. It has been
assumed in this study that unsupported panels 10 mm thick can be
encountered in the museum collections and, therefore, a realistic
objects are usually extensively cracked. The cracking has been
thickness of panels ranges between 10 and 40 mm. The times of
considered an ‘acclimatisation’ process for painted wood with
the 95% moisture response for panels 10 and 40 mm thick with
cracks acting as expansion joints and thus reducing the risk of
both faces open to the moisture flow – to illustrate the worst case
further mechanical damage from fluctuations which do not go
of design layers providing little barrier effect to water vapour
beyond the past pattern [13].
diffusion – was determined as 6 and 100 days, respectively [17].
Since the predicted RH values taken from the Hadley Model are
4.1.2. Translating critical elongations to critical magnitude of the used in the form of the 30-day averages, they will bring about a full
relative humidity variations moisture response of thinner panels encountered in the collections
The moisture coefficients of dimensional change for wood sub- and produce the damaging impact on the gesso predicted by
strate and gesso are required to translate the critical elongation the S-N curve of Fig. 2. Such relatively thin, moisture-responsive
values of the S-N curve into the critical magnitudes of RH variations panels are regarded in this study as the worst-case category of
necessary to cause these elongations. objects on which the assessment of the damage function, relating
The moisture coefficient of dimensional change for wood was environmental impacts to panel painting damage, is based.
obtained from the ‘general’ swelling and shrinkage isotherms
obtained by fitting the swelling and shrinkage data measured 4.2. Transfer functions to predict indoor climate from the outdoor
at 24 ◦ C for 21 wood species widely used in the past for panel climate data
paintings and woodcarving [7]. The isotherms obtained, shown for
the tangential direction in Fig. 3, can be regarded representative, for As historic painted wooden objects are preserved indoors,
most practical purposes, of mixed wood collections or furnishings. indoor climate needs to be predicted from future estimates
A distinctive and reproducible hysteresis loop appears between of outdoor climate, available from the climate change models.
the two branches: the swelling branch measured on increasing Transfer functions predicting indoor climatic parameters from
RH from the dry state and the shrinkage branch measured on outdoor climate were, therefore, derived for unheated buildings
decreasing RH from a moist state. The hysteresis leads to higher by correlating the measured outdoor and indoor climatic data.
moisture-induced elongation during desorption when compared The same general approach has been recently used by Lankester
to that during adsorption at any given RH value. The initial elon- and Brimblecombe [18]. As heating can introduce serious destabi-
gation of wood at 50% RH was assumed to be an outcome of the lization to the natural indoor climate, which depends on selected
swelling, shrinkage and swelling path following the changes in heating strategies and systems, analysis of transfer functions
RH: 0% →96% →31.5% which had been selected to ensure that this predicting conditions in heated interiors is a separate issue, which
initial elongation was the arithmetical average of elongation values is outside the scope of this paper.
corresponding to 50% RH on the swelling and shrinking branches. Indoor climate in unheated historic buildings is essentially
The intermediate – or scanning – swelling/shrinking curves were governed by the outdoor climate, the properties of the building
generated from that starting elongation using the approach envelope and the building usage patterns, for example opening
described in Ref. [14], and used to determine the dimensional of doors and windows or presence of visitors. Generally, the
change of wood in the tangential direction when RH fluctuated ventilation in historic buildings is natural with the ventilation
between 65 and 35%, that is, ± 15% around the mid-value of 50% rates varying widely between 0.1 to 3 air changes per hour (ACH)
RH. The entire procedure is illustrated in Fig. 3. depending on the gaps in building structure, tightness of windows
The moisture coefficient of dimensional change for gesso at PVC and doors, pressure differences induced by wind and temperature,
of 92% was calculated as 9.4 × 10−6 per 1% of RH using the formula and the pattern of use. However, it is known that a substantial
derived by Michalski [8] assuming that the moisture coefficient of number of historic buildings such as churches and castles have
dimensional change for the pure glue is 4 × 10−4 per 1% of RH. The high ventilation rates, ranging from 0.5–0.75 ACH for churches
value calculated shows an excellent agreement with values pro- with wooden vaults [19] to 3 ACH observed in wooden historic
vided by Mecklenburg [15]. High concentration of the inert filler structures [20]. In this paper, the transfer functions between indoor
368 Ł. Bratasz et al. / Journal of Cultural Heritage 13 (2012) 365–370

and outdoor parameters are derived for such buildings of high

ventilation rates regarded as a worst-case category. It is assumed
that a high air exchange between the interior and the outside
results in a very close moisture contents in outdoor and indoor
air, for example expressed as humidity mixing ratio (MR), that
is, grams of water vapour per kilogramme of dry air. In turn, the
variation in the temperature indoors are evened out and shifted in
time when compared with outdoors due to impeded air exchange
and the buffering effect of the building. The variations in RH can
be calculated from the known MR and the temperature in the
building. Indoor climate in one of the rooms in the Teutonic Knights
Castle in Malbork – World Heritage Site, Poland has been chosen
as representative of an unheated interior to calculate the transfer
function. The measured ventilation rate was 0.5 ACH and the MR
values measured over the period of 18 months in the outdoor and
indoor air was practically identical – the linear regression of the
MRoutdoor vs. MRindoor plot yielded the intercept and the slope of
Fig. 5. Indoor relative humidity (RH) predicted for the year 2070 from the HadCM3
0.154 and 0.984 respectively, with the determination coefficient R2 climate model. Plot of 30-day averages (upper part) is compared with the set of
of 0.997. The relationship between the indoor tindoor and outdoor simple cycles (bottom part) obtained by the rainflow analysis.
toutdoor temperatures, expressed in ◦ C, was linear and had a form
tindoor = 2.9 + 0.94 toutdoor , the determination coefficient R2 being
4.3. Climatological risk index to painted wood
0.973. The upward and downward deviations of the measured
values from the average linear plot reflect thermal inertia of the
The real-world RH variations, measured or predicted by the
room: after winter the cold walls keep the indoor temperature
climate change models, are a combination of cycles of various dura-
lower than outdoors and vice versa in autumn and winter, the
tion and amplitudes. By way of example, a plot of the predicted
room stays warmer than outdoors. The indoor air temperature
30-day average RH values in a year 2070 for the HadCM3 Model
follows the outdoor tendency with a delay of approximately four
cell in which Paris is located, is shown in Fig. 5. The variations
days as revealed by comparing the plots of the indoor tempera-
can be reduced mathematically into simple RH cycles of various
tures measured and calculated (Fig. 4). The magnitude of the delay
sizes using the rainflow counting method [22]. The outcome of the
was determined using least square fitting technique to minimize
analysis is shown in the bottom part of Fig. 5. The detailed proce-
difference between the measured and calculated values. Finally,
dure is described in the reference and is not covered here but it can
indoor RH was calculated from MR and temperature indoors
be seen how the method takes into account RH cycles of various
[21]:
duration embedded in the complex climate change prediction. The
yearly RH cycle is counted as two one-half cycles corresponding to a
RHindoor (tindoor corr ) = 26.35 ∗ MR ∗ 10a∗tindoor corr /(b+tindoor corr )
decrease by 34% RH from the largest peak in January to the smallest
valley in July and to an increase by 36% RH from the same smallest
where a = 7.5, b = 237.3 ◦ C for tindoor corr ≥ 0 and a = 9.5, b = 265.5 ◦ C valley in July to the largest peak in December, respectively. Two full
for tindoor corr < 0, tindoor corr is tindoor shifted by 4 days to get rid of the smaller cycles of ranges not exceeding 15% RH are marked at their
thermal inertia effect. The RH values calculated by the described starting points in March and September. The plot of the predicted
procedure and measured are compared in Fig. 4. The correlation 30-day average RH values for the same area but covering the entire
coefficient is 0.93 indicating that much of the observed varia- period of the far future i.e. between 2070 and 2099, together with
tions in the RHindoor were accounted for by the transfer function the outcome of the rainflow counting of the single cycles, are illus-
adopted. trated in Fig. 6. The same analyses were carried out for two other
periods considered: the baseline and the near future covering the
periods of 1961–1990 and 2010–2039, respectively. The results are

Fig. 4. Indoor temperature and relative humidity (RH) in a room in the Teutonic
Knights Castle in Malbork, Poland during period January 2006 to July 2007. Measured Fig. 6. Indoor relative humidity (RH) predicted for the period of 2070–2099 from
and predicted 30-day central moving averages calculated of the data points recorded the HadCM3 climate model. Plot of 30-day averages (upper part) is compared with
every five minutes in the two adjacent 15-day periods. the set of simple cycles (bottom part) obtained by the rainflow analysis.
 Ł. Bratasz et al. / Journal of Cultural Heritage 13 (2012) 365–370 369

Fig. 8. Climatological risk index to painted wood mapped over Europe for
2070–2099 derived from HadRM3 output under the A2 scenario.
Fig. 7. Histogram showing the number of times single cycles of various sizes occur.
Outcome of the rainflow analysis of time-series of 30-day relative humidity (RH)
averages for three periods indicated.

summarised in a histogram providing the number of times cycles

of various sizes occur in each of the periods analysed (Fig. 7). It is
noticeable that large RH cycles of ranges exceeding 25% RH are pre-
dicted to increase in the future, the tendency especially evident for
the far future.
Thirty days exceed considerably the response time of 10 mm
painted panels identified in Section 4.1.2 as the worst-case category
particularly vulnerable to climate-induced damage. Therefore, the
equilibrium moisture content in the panels and the related elonga-
tion will follow continuously the changes in the 30-day average RH
values. As a result, the climatological risk index C for painted wood
is defined as:
⎛ ⎞

k
n 
l
Aj
C=⎝ i
+ ni ⎠× 1
Ni A0 a
i=1 j=1

where k is number of RH ranges smaller than A0 = 12% RH (a thresh-

old range for fracture of the design layer in a single cycle) - each
contributing ni cycles and Ni being number of cycles to failure for
a given RH range; j is number of RH ranges greater than A0 = 12%
RH - each contributing nj cycles of range Aj . a is number of years
over which the RH cycles are counted from the irregular RH-versus-
time predictions using the procedure described in this paper. The
first component in the brackets, known in the materials science
as Miner’s rule, assesses the combined damaging effect of the
below-threshold RH cycles on the design layer. The second com-
ponent assesses the combined damaging of the above-thresholds
RH cycles, the damaging impact of each cycle range being weighted
Fig. 9. Difference maps from baseline (1961–1990) to near future (2010–2039) and
by the ratio of a given range Aj to the threshold range A0 . The index,
far future (2070–2099) representing the changes in the climatological risk index to
therefore, is calculated as a yearly ‘dose’ or accumulated fatigue painted wood derived from HadCM3 output under the A2 scenario.
damage of the pictorial layer from the real-world climatic vari-
ations measured or predicted. It should be recalled at this point
that the risk index is very conservative as the damaging levels of patterns of risk loosely follow the predicted increases in rain days
elongation – on which the index is based – are assumed to cause across Europe from the HadCM3A2a model run [3].
the first incidence of fracturing on undamaged gesso. Usual, exten-
sive cracking of historic design layers reduces the risk of further 5. Conclusions
mechanical damage with cracks acting as expansion joints when
the design layer is elongated. The experimental approach, involving the cyclic mechanical
The index has been used to produce maps illustrating the stretching of specimens imitating historic panel paintings and the
predicted risk of damage to painted wood across Europe. Fig. 8 illus- monitoring of the development of cracks in the design layer on
trates the index for the far future obtained with the high-resolution wood, has allowed the determination of critical levels of elonga-
HadRM3 regional climate model. Areas, which may be significantly tion, above which damage in this layer appears, as a function of
affected by climate change, are highlighted on the difference map the number of elongation cycles. The critical magnitudes of RH
from baseline to the near and far future, respectively (Fig. 9). The variations necessary to cause these elongations were derived. The
370 Ł. Bratasz et al. / Journal of Cultural Heritage 13 (2012) 365–370

detailed insight into the damage potential of the RH variations [4] M.F. Mecklenburg, C.S. Tumosa, D. Erhardt, Structural response of painted wood
allowed the definition of a climatological risk index based on the surfaces to changes in ambient relative humidity, in: V. Dorge, F.C. Howlett
(Eds.), Painted Wood: History and Conservation, The Getty Conservation
number and the magnitude of the variations of the 30-day RH Institute, Los Angeles, 1998, pp. 464–483.
averages. The index is calculated as a yearly ‘dose’ or accumulated [5] R. Kozłowski, Ł. Bratasz, Ł. Lasyk, M. Łukomski, Allowable microclimatic vari-
fatigue damage of the pictorial layer from the real-world climatic ations for painted wood: direct tracing of damage development, in: A.S.
Chui, A. Phenix (Eds.), Proceedings of Symposium ‘Facing the Challenges of
variations measured or predicted by the climate change scenarios. Panel Paintings Conservation: Trends, Treatments and Training’, Los Ange-
As historic painted wooden objects are preserved indoors, transfer les, May 17–18 2009, The Getty Conservation Institute, Los Angeles, 2011,
functions predicting indoor climatic parameters from outdoor pp. 158–164.
[6] Ł. Lasyk, M. Łukomski, Ł. Bratasz, Simple DSPI for Investigation of Art Objects,
climate data were derived for unheated historic buildings. The
Optica Applicata 41 (2011) 687–700.
output from climate models can then be used to predict the future [7] Ł. Bratasz, R. Kozłowski, A. Kozłowska, B. Rachwał, Sorption of moisture and
threat to painted wood indoors. dimensional change of wood species used in historic objects, in: J. Gril (Ed.),
Proceedings of the International Conference held by COST Action IE0601 ‘Wood
The presented methodology is not limited to forecasting the
Science for Conservation of Cultural Heritage’, Braga, November 5–7 2008,
damaging impact of future outdoor climate change on painted Firenze University Press, Florence, 2010, pp. 11–16.
wood. It can be also used to assess quality of climate stabilisation [8] S. Michalski, Crack mechanism in gilding, in: D. Bigelow, E. Come, G.J. Landrey,
in more precise way than offered by general climate classes like C. Van Horne (Eds.), Gilded Wood: Conservation and History, Sound View Press,
Madison, Connecticut, 1991, pp. 171–181.
ones proposed by ASHRAE [17]. The analysis can be applied to [9] T.C. Johns, J.M. Gregory, W.J. Ingram, C.E. Johnson, A. Jones, J.A. Lowe,
both natural climates in buildings with no active climate-control J.F.B. Mitchell, D.L. Roberts, D.M.H. Sexton, D.S. Stevenson, S.F.B. Tett, M.J.
systems and to artificial climates in buildings where active systems Woodage, Anthropogenic climate change for 1860 to 2100 simulated with
the HadCM3 model under updated emissions scenarios, Clim. Dyn. 20 (2003)
operate to control air temperature (heating), air temperature and 583–612.
humidity (air-conditioning), or ventilation rate. Similar indices [10] C.C. Gordon, C.A. Senior Cooper, H.T. Banks, J.M. Gregory, T.C. Johns, J.F.B.
were derived to assess climate quality in respect to risk of chemical Mitchell, R.A. Wood, The simulation of SST, sea ice extents and ocean heat trans-
ports in a version of the Hadley Centre coupled model without flux adjustments,
degradation or mould growth in cultural heritage institutions. Clim. Dyn. 16 (2000) 147–168.
Interesting examples of such indices are the isoperm value and the [11] V. Pope, M.L. Gallani, P.R. Rowntree, R.A. Stratton, The impact of new physical
mould growth hazard index used by the National Archives UK to parameterizations in the Hadley Centre climate model: HadAM3, Clim. Dyn. 16
(2000) 123–146.
evaluate impact of the developed environmental strategies on the
[12] IPCC (Intergovernmental Panel on Climate Change). IPCC Special Report Emis-
stored documents [23]. Another potential application of the risk sion Scenarios: Summary for Policymakers, 2000.
indices is evaluation of the harmlessness, or otherwise, of various [13] S. Michalski, The ideal climate, risk management, the ASHRAE chapter, proofed
fluctuations, and towards a full risk analysis model, in: F. Boersma (Ed.),
strategies for reducing the environmental control needed to ensure
Proceedings of Experts’ Roundtable on Sustainable Climate Management
good preservation of the objects, using less and simpler equipment Strategies, Tenerife Island, April 2007. The Getty Conservation Institute,
and lowering investment, maintenance and energy costs. In this Los Angeles, http://getty.edu/conservation/science/climate/climate experts
way, low-energy solutions can be safely introduced to make the roundtable.html. (Accessed 28 November 2011).
[14] S. Jakieła, Ł. Bratasz, R. Kozłowski, Numerical modelling of moisture movement
historic buildings sustainable so that even as the energy cost rise and related stress field in lime wood subjected to changing climate conditions,
and limits on carbon emissions are imposed, high standards of Wood Sci. Technol. 42 (2008) 21–37.
preservation are maintained with affordable and efficient means. [15] M.F. Mecklenburg, Some mechanical and physical properties of gilding gesso,
in: D. Bigelow, E. Come, G.J. Landrey, C. Van Horne (Eds.), Gilded Wood: Con-
servation and History, Sound View Press, Madison, Connecticut, 1991, pp.
Acknowledgements 163–170.
[16] L. Uzielli, Historical overview of panel-making techniques in Central Italy, in:
K. Dardes, A. Rothe (Eds.), The Structural Conservation of Panel Paintings, The
This research was supported by grant PL0086 from Iceland, Getty Conservation Institute, Los Angeles, 1998, pp. 110–135.
Liechtenstein and Norway through the European Economy Area [17] Museums, Galleries, Archives and Libraries, Chapter 21, in: ASHRAE Handbook
Financial Mechanism and a grant from the Polish Ministry of Sci- – HVAC applications, ASHRAE – American Society of Heating, Refrigerating, and
Air-Conditioning Engineers Inc. (2007).
ence and Higher Education supporting activities of COST Action [18] P. Lankester, P. Brimblecombe, Future thermohygrometric climate within his-
IE0601 “Wood science for conservation of cultural heritage”. toric houses, J. Cult. Herit., 2011, doi:10.1016/j.culher.2011.06.001.
[19] H. Schellen, Heating Monumental Churches, Indoor Climate and Preser-
vation of Cultural Heritage, Eindhoven Technical University, Eindhoven,
References 2002.
[20] L. Samek, A. De Maeyer-Worobiec, Z. Spolnik, L. Bencs, V. Kontozova, Ł. Bratasz,
[1] Climate Change and Museum Collections, a discussion held on September R. Kozłowski, R. Van Grieken, The impact of electric overhead radiant heat-
17, 2008 London, edited transcription, International Institute of Conser- ing on the indoor environment of historic churches, J. Cult. Herit. 7 (2007)
vation, www.iiconservation.org/dialogues/IIC climate change transcript.pdf. 361–369.
(Accessed 28 November 2011). [21] D. Camuffo, Microclimate for Cultural Heritage, Developments in Atmospheric
[2] The Plus/Minus Dilemma: The Way Forward in Environmental Guide- Science 23, Elsevier, Amsterdam, 1998.
lines, a discussion held on May 13, 2010 Milwaukee Wisconsin, USA, [22] ASTM Standard E 1049–85, Standard practices for cycle counting in fatigue
edited transcription, International Institute of Conservation, www.iiconser- analysis, 2005.
vation.org/dialogues/Plus Minus trans.pdf. (Accessed 28 November 2011). [23] S. Hong, M. Strlič, I. Ridley, K. Ntanos, N. Bell, M. Cassar, Monitoring and mod-
[3] C. Sabbioni, P. Brimblecombe, M. Cassar, The Atlas of Climate Change Impact elling the storage environment at The National Archives, United Kingdom,
on Cultural Heritage: Scientific Analysis and Management Strategies, Anthem in: J. Bridgland (Ed.), ICOM Committee for Conservation, 16th Triennial
Press, London, 2010. Conference, Lisbon: Preprints, 2011.