Interactive comment on “ European summer climate modulated by NAO-related precipitation ”

1. On general grounds the statement that a statistical technique can determine the direction of influences cannot be true. If all the information available is two covarying fields, no statistical technique can ever show causality. In particular, one can never exclude that a third factor influences both fields. As an example, El Niño causes both drought in September–October in Indonesia, and often fewer C2459


Introduction
The recent European climate is characterized by an increasing frequency of summer heat waves with substantial societical and ecological impacts, e.g. the recordbreaking heat wave in 2003.Climate projections point towards even higher-frequent Introduction

Conclusions References
Tables Figures

Back Close
Full and longer-lasting heat waves under increased greenhouse gas emission scenarios (Scherrer et al., 2005;Pal et al., 2004;Stott et al., 2004;Meehl et al., 2004).These past and projected heat waves highlighted the importance of a detailed understanding of the mechanisms that contribute to the initialization and persistence of extreme heat conditions.Hot and dry summers in Europe are generally associated with a specific, large-scale anticyclonic atmosphere circulation regime (Cassou et al., 2005;Fischer et al., 2007).Most of hot and dry summers over Europe were preceded by pronounced deficits of precipitation in winter and early spring (Della-Marta et al., 2007;Vautard et al., 2007).Vautard et al. (2007) showed with the mesoscale MM5 model that the observed winter precipitation deficit and summer heat wave were dynamically linked via land-atmosphere feedback loops, wherein soil moisture played a crucial role.The deficit of precipitation and subsequent drier soils resulted in reduced latent cooling and thereby an increase of air temperature, in agreement with other numerical experiments (e.g., Seneviratne et al., 2006;Fischer et al., 2007;Zampieri et al., 2009).These investigations of individual heat waves highlighted the role of land-atmosphere interactions particularly between winter precipitation and subsequent soil moisture states but also pointed to the importance of circulation patterns in the generation of summer heat waves.An immediate question that arises is whether this land surface feedback mechanism exists only for extraordinary hot summers or more systematically.Sch är et al. (2004) underlined that an increase of interannual temperature variability in response to greenhouse-gas forcing might be an alternative causal mechanism for the occurrence of European summer heat waves; and numerical analysis with prescribed soil moisture by Seneviratne et al. (2006) suggested further that the increased interannual temperature variability is strongly related to the land-atmosphere interactions.However, there exists as yet no clear analysis of observational evidence connecting summer temperature to interannual variability of winter precipitation.The present paper aims to fill this gap in our understanding by investigating with long-term observations the summer temperature variability, including mean and maximum, in relation to the interannual variability of winter precipitation and possible relevant circulation regimes.Introduction

Conclusions References
Tables Figures

Back Close
Full Furthermore, we use a soil moisture proxy to clarify if these responses are related to soil moisture processes.The paper is organized as follows: in Sect. 2 the observational datasets used are described and the analysis technique is briefly introduced.Section 3 is dedicated to the results and, finally, Sect. 4 contains a discussion and the conclusions of this study.

Datasets
We use long-term gridded observations of accumulated precipitation in January-March (P JFM ) and averaged daily mean as well as maximum temperature in summer (June-August, T mean and T max , respectively) for the period 1901-2005, derived from University of East Anglia Climatic Research Unit (CRU) at a horizontal resolution of 0.5 • ×0.5 • (Mitchell et al., 2005).The P JFM values over mountain Scandinavia are not included in this study.Due to the sparseness of in situ soil moisture observations, the averaged self-calibrating Palmer drought severity index in June-August (scPDSI, Wells et al., 2004) is used as a proxy of soil moisture.

Coupled manifold technique
Widely used methods to detect coupling between climatic fields are variance analysis methods, such as Maximum Covariance Analysis (MCA, also termed as SVD) to Introduction

Conclusions References
Tables Figures

Back Close
Full maximize explained covariance or canonical correlation analysis (CCA) to maximize correlations.These traditional methods require orthogonal solutions with little physical justification, and can detect unfortunately only cross-correlation between fields.However, the climate system is full of interactions and a cross-correlation provides little information of directional influences between fields.Recently, Navarra and Tribbia (2005) proposed a non-orthogonal solution for this problem, the Coupled Manifold Technique (CMT), which enables to detect the directional interactions.It is mathematically demonstrated that the CMT technique provides a framework that generalizes the traditional methods for variance analysis, such as SVD, CCA and regression analysis.
Assuming that the relation between two given fields Z and S is linear, the CMT looks for the response of Z to S using a linear operator A that satisfies where the norm is the Frobenius norm, defined by and the primes denotes a matrix transpose operation.Using the Procrustes method (Richman et al., 1993), the A solution can be written as The operator A represents the influence of S on Z.Using A we can separate the field Z into two parts: The Z for part is the portion of the field variability that is forced by the S variability (henceforth "forced manifold"), while Z free is the portion independent from S ("free manifold").
In the same way, that part of S field forced by Z variability can be isolated by reversing the input-output roles of the two fields.Since the direction of influence between Introduction

Conclusions References
Tables Figures

Back Close
Full fields is taken into account, the CMT technique detects causality rather than crosscorrelation, and thus constitutes a significant improvement over traditional SVD and CCA techniques.The CMT has been used to study directional interactions between vegetation and atmosphere, showing good performance (Alessandri et al., 2008).To simplify the computation, this technique is applied to the EOF coefficients of fields of interest in our study, as suggested by Navarra and Tribbia (2005), with 99% of the total variance of each field retained.Furthermore, each element of A is tested against the null hypothesis of being equal to zero at the 1% significance level based on the Student t distribution as described by Cherchi et al. (2007).This method may identify both one-way and two-way relations between fields.In our analysis, there exist time lags between fields, and therefore we end up with only one-way relation, that is, only the variability in Z forced by S. When the forced manifold is obtained, a further significance test of the forced variance is performed to make the result robust using a Monte Carlo approach.Then S and the forced manifold Z for , containing now only the variability in Z forced by S, are subjected to traditional MCA to obtain the forcing and forced patterns as well as time coefficient series of interest.

Responses of T mean and T max to P JFM
Figure 1a shows the percentage of T mean variance forced by the P JFM variability, and that for T max is shown in Fig. 1d.These values are derived from the ratio of the forced T mean (T max ) manifold to the original T mean (T max ) fields.We tested where the percentage values are significant different from zero at the 0.10 level.For each grid point, we tested the null hypothesis of getting as high or higher variance fractions through a Monte Carlo bootstrap method (10 000 repetitions of the CMT) by randomizing the order number of P JFM values on each grid.The largest values are found over Southern Europe for both T mean and T max where it is most sensitive to land-atmosphere interactions (Seneviratne Introduction

Conclusions References
Tables Figures

Back Close
Full  , 2006;Fischer et al., 2007;Zampieri et al., 2009) The MCA analysis was originally designed for detecting cross-correlation.In our study it is conducted to the P JFM field and the T mean (T max ) manifold forced by the P JFM variability, and therefore what it detects is the forcing P JFM pattern and the T mean (T max ) response.Derived from the first MCA mode, Fig. 1b and c shows the 1st pair of forcing P JFM pattern and its T mean response, containing 95% of the total squared covariance.This MCA mode exhibits unit correlated time coefficients (r>0.999),suggesting the derived forcing-forced relationship is very robust.The time coefficient series are shown as blue lines in Fig. 3.We note that the unit correlation derived here is due to the data preprocessing with CMT, which constructs only the T mean variability forced by P JFM at significance level of 0.01.The time coefficient series of the 1st MCA mode without CMT exhibit a correlation of 0.40 (not shown), which is clearly insufficient to conclude a significant linkage.The same situation also holds in the following analysis of T max as well as soil moisture proxy of scPDSI.
Shown in Fig. 1b and c, there exists only one significant P JFM anomaly over the Mediterranean, with opposite sign of the T mean response largely northward and eastward extended to 50 • N compared to the P JFM anomaly.This suggests that T mean in summer fluctuates in correspondence to the anomalous states of P JFM via the cooling effect of the surface energy balance.Precipitation is spectrally white with very limited Introduction

Conclusions References
Tables Figures

Back Close
Full memory up to two weeks due to the chaotic nature of atmosphere (Wang et al., 2010); therefore the extended memory of P JFM is probably sustained by soil moisture feedbacks on precipitation.One may question the existence of forced T mean anomalies in the opposite sign over north of 50 • N. It appears not a physical response to P JFM since there exists no forcing anomaly in P JFM in the very location.Therefore we attribute it to be a statistical coexistence with no physical implication.Furthermore, this anomaly accounts for a very low percentage of the forced T mean variance over north of 50 • N (Fig. 1a) and did not pass our significance test.Shown in Fig. 1e and f is the 1st leading pair of the forcing P JFM pattern and its T max response, which contains 96% of the total squared covariance with unit correlated time coefficient series (r>0.999,blue lines in Fig. 3).Comparing Fig. 1b and e, we can see clearly the T max anomaly is forced by almost the same P JFM anomaly as that forces T mean .Furthermore, these time coefficient series are nearly unit correlated with those derived from the P JFM ∼T mean association (r>0.999).These statistical properties suggest that the derived linkages between P JFM and T mean as well as T max are very likely to be driven by the same climate dynamics.The forced T max (Fig. 1f) exhibits very similar dipole pattern as T mean (Fig. 1c), however, the percentage of forced variance over north of 50 • N is small again.
An important question regarding land-climate interactions is whether they lead to amplified variability of climate extremes, such as heat waves, particularly in the context of climate change (Seneviratne et al., 2010).Over south of 50 • N, the percentage of T max variance forced by P JFM appears to be more homogenized than that of T mean .Furthermore, the robust relations derived from MCA analysis after CMT enable us to compare the magnitudes of T mean and T max responses to P JFM , where the magnitude of the T max response appears to twice that of T mean .Therefore P JFM exerts to some extend larger influence on T max than that on T mean over south of 50 • N, possibly through water cycle interactions.Introduction

Conclusions References
Tables Figures

Back Close
Full

The role of soil moisture
So far we have shown that summer temperature fluctuates in relation to fluctuations in winter precipitation over Mediterranean.It is plausible to hypothesize that these responses are modulated by interactions between the water cycle and temperature with soil moisture playing a critical mediating role.An analysis of soil moisture would help to support this hypothesis.For this purpose the same analytic framework as above is conducted to P JFM and summer scPDSI as a soil moisture proxy.Analysis of this field is expected to clarify the role of soil moisture in the forced T max and T mean responses to P JFM .This analysis is restricted to south of 55 • N where P JFM has distinctive expressions in the T max and T mean fields.
The scPDSI variability forced by P JFM is shown in Fig. 2. Shaded values in Fig. 2a indicate the percentage of scPDSI variance forced by P JFM that can pass significance test at 0.01 level, with the largest values of 20∼25% existing in the West Mediterranean.The 1st MCA mode contains 80% of the total square covariance with unit correlated time coefficient series (r>0.999,green lines in Fig. 3).The forcing P JFM pattern exhibits a distinctive anomaly over Mediterranean (Fig. 2b), very similar to that P JFM patterns forcing T mean and T max (Fig. 1a, d).The scPDSI response is of the same sign but largely northward and eastward extended (Fig. 2c) compared to the forcing P JFM pattern.Of particular interest is that the time coefficient series are highly correlated with those from temperature analyses in Sect.3.1, with correlation coefficient r>0.999.The MCA analyses are summarized in Table 1.Therefore the responses of scPDSI, T mean and T max to the P JFM variability we present here appear to be driven by the same climate dynamics, and P JFM is very likely to influence T mean and T max via soil moisture.That is, a negative precipitation anomaly in winter is supposed to result in summer heating due to reduced latent cooling from soil moisture.The reverse relationship also holds, where a positive precipitation anomaly implies cooling.
These observational relations corroborate the interactions between water cycle and temperature established in previous numerical work, e.g., Seneviratne et al. (2006).Introduction

Conclusions References
Tables Figures

Back Close
Full Note that if we perform the same set of statistics to the winter precipitation and summer minimum temperature, we do not obtain the same relations.This is physically reasonable because the minimum temperature is highly constrained by external forcings, such as atmospheric circulation and sea surface temperature, rather than internal feedbacks (Alfaro et al., 2006;Zhang et al., 2008).

Link to North Atlantic Oscillation
The North Atlantic Oscillation (NAO) is the dominating large-scale atmospheric circulation over the Atlantic-Europe sector in winter, with marked influence on winter climate.
In recent years, the NAO is also observed to influence summer climate over Europe, in a weak but significant way.For example, Qian et al. (2003) showed with observations that European summer temperature has positive correlation with the NAO index in previous January and February; Kettlewell et al. (2003) discovered a negative correlation between winter NAO and summer precipitation over Europe.However, the mechanism that links these phenomena remains still a puzzle.The derived forcing P JFM patterns on T mean and T max as well as scPDSI in our analysis appear to resemble the NAO regime over Mediterranean, suggesting a plausible hypothesis that the NAO variability modulates summer climate over Europe through controlling winter precipitation that subsequently initializes the moisture states of water cycle interactions with temperature.
To further clarify the role of NAO in these processes, we compared the winter NAO into Europe result in a dry Mediterranean Europe during a high NAO winter and the opposite during a low NAO winter (Hurrell et al., 2001).Based on the above analysis, we suggest that the NAO regime over the Mediterranean modulates European summer climate via initialization of the winter land surface moisture.

Discussion and conclusion
The importance of soil moisture initialization in winter and early spring for the seasonal prediction of heat and drought waves in European summer has been demonstrated in recent years (e.g., Vautard et al., 2007;Fischer et al., 2007;Zampieri et al., 2009;Seneviratne et al., 2006;Ferranti et al., 2006).Although soil moisture is closely related to precipitation, a clear picture of the relations between summer climate and preceding winter precipitation has not yet been demonstrated observationally.This is largely because the expected signal is very weak in the fields of interest, and traditional techniques for cross-correlation, such as MCA and CCA, are not capable of generating robust relations from this strong background noise.
Using the newly developed CMT technique that detects directional influence between climatic fields, we present in this paper robust responses of summer T mean and T max as well as scPDSI to previous winter precipitation.Distinctive responses exist only over the Mediterranean area, where the temperature response is most sensitive to landatmosphere interactions in regional climate models (Sch är et al., 1999;Seneviratne et al., 2006).The P JFM variability accounts for up to 10∼15% of the total T mean and T max variance, respectively for the period of 1901-2005; for the scPDSI this value amounts to 20∼25% over the Western Mediterranean.The P JFM appears to influence T mean and T max via scPDSI, agreeing very well with our recent understanding of the water cycle dynamics over land (see Seneviratne et al., 2010 for a review).Therefore our findings are very likely to be physical of origin, although there is always a risk to infer physics from statistics.We also note that we are not addressing the full picture of landatmosphere feedback processes but only that part that is related to January-March precipitation.Figures

Back Close
Full The extension of responses towards north and east is also observed in numerical experiments.Vautard et al. (2007) suggested that the northward propagation may be due to the southerly wind episodes carrying moisture northward.The eastward propagation is probably due to the heat low response over Central Europe, blocking the inflow of moist maritime air from the Atlantic and reinforcing the northward extension dynamically, addressed by Haarsma et al. (2008).Using a moisture tracer model, Bisselink and Dolman (2009) also found that advection is the most important contributor to precipitation over Central Europe.It is notable that the largest anomalies of T mean and T max responses to P JFM (Fig. 1c, f) exist in Central Europe, while the largest T mean and T max variance forced by P JFM exists in Southeast Europe (Fig. 1a, d).This is because the interannual variability of T mean as well as T max over Central Europe is much stronger than that over Southeast Europe.We infer that the forced variance can be considered as the forcing strength or land-atmosphere coupling strength, while the forced anomalies cannot.It can also be noticed that the P JFM variability forces large T mean and T max variances over Southeast Europe (Fig. 1a, d), but that the variance of scPDSI is very limited there (Fig. 2a).This is possibly because the derived T mean and T max variability over Southeast Europe is closely related to the eastward extended heat response (Haarsma et al., 2008), while the soil moisture availability is very small there (Bisselink and Dolman, 2009).
We suggest that the NAO regime over the Mediterranean modulates summer climate over Europe through controlling winter precipitation that then initializes water cycle interactions with temperature.A positive phase of NAO tends to cause a hot and dry summer, or vice versa.This suggests there is scope for improved seasonal prediction of heat and drought waves from the pressure pattern of winter NAO.A remarkable feature of the NAO is its prolonged positive phases in the past 40 years, possibly related to anthropogenic warming (Shindell et al., 1999).This NAO dry pattern over the Mediterranean may have contributed to the increased frequency of heat and drought waves since then through modulating the water interactions over the Mediterranean.Introduction

Conclusions References
Tables Figures

Back Close
Full  Full Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | et al.
Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | index and the time coefficient series derived from MCA analysis.We use the averaged values of NAO index in January-March for the period 1901-2005, based on the difference of normalized sea level pressures between Gibraltar, the Azores and SW Iceland.Shown in Fig. 3, the derived time coefficient series from T mean , T max and scPDSI analyses have a high correlation with the NAO index with r=0.65 (p<0.05),suggesting a significant relation between the NAO variability and summer climate.The NAO variability is a north-south shift (or vice versa) in the track of storms and depressions across the North Atlantic Ocean and into Europe.The Atlantic storms that travel Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper |

Fig. 1 .
Fig. 1.T mean as well as T max variability forced by P JFM .(a) Percentage of T mean variance forced by P JFM (sig=0.10 in the red rectangle).(b) The forcing P JFM pattern and (c) its T mean response.(d) Percentage of T max variance forced by P JFM (sig=0.10 in the red rectangle).(e) The forcing P JFM pattern and (f) its T max response.All the relevant time coefficient series mutually exhibit unit correlation (r>0.999),shown in Fig. 3. Units are K for T mean as well as T max and mm for P JFM .

rather complex water budget model involving water cycle interactions with tem- perature; therefore it is suitable for the purpose of this study. Ideally one would use remotely sensed soil moisture observations
(e.g., de Jeu et al., 2008)but the datasets are unfortunately not yet sufficiently long in time.The scPDSI dataset obtained from CRU spans 1901-2002 on a monthly basis and range from −4 to +4 in the case of extremely dry and extremely wet conditions, respectively (van der Schrier et al., 2006).
The scPDSI is based on soil water content in a , while little forcing (low values) is observed over Northern Europe of 50 • N. Up to 5∼15% of the summer T mean variance over Southern Europe appears to be forced by P JFM .The forced T mean variance by P JFM is up to 8% over Western Europe, averaged within the green rectangle in Fig. 1a, which doesn't pass the significance test.Over Eastern Europe, this value increases to 11% averaged within the red rectangle in Fig. 1a, passing the significance test.This implies that summer T mean over Eastern Europe is more sensitive to P JFM .These values for T max are a bit higher.The forced T max variance is up to 10% over Western Europe and that value over Eastern Europe is up to 14%, averaged within the green as well as the red rectangles, respectively in Fig. 1d.Low values for both T mean and T max over North of 50 • N indicate little influence from P JFM .

Table 1 .
A summery of the MCA analyses between P JFM and the forced manifolds.