Abstract
In this paper, we propose a novel approach for generating avalanche hazard maps based on the spatial dependence of avalanche runout altitudes. The right-truncated data are described with a Bayesian hierarchical model in which the spatio-temporal process is assumed to be the sum of independent spatial and temporal terms. Topography is roughly taken into account according to valley altitude and path exposition, and the spatial dependence is modelled with a Matérn covariance function. An application is performed to the Haute-Savoie region, French Alps. A spatial dependence in runout altitudes is identified, and an effective range of about 10 km is inferred. The temporal trend extracted highlights the increase of avalanche runout altitudes from 1955, attributed to both anthropogenic factors and climate warming. In a cross validation scheme, spatial predictions are provided on undocumented paths using kriging equations. All in all, although our model is unable to take into account small topographic features, it is a first-ever approach that produces very encouraging results. It could be enhanced in future work by incorporating a numerical physically-based code into the modelling.
Similar content being viewed by others
References
A’Hearn B (2004) A restricted maximum likelihood estimator for truncated height samples. Econ Human Biol 2(1):5–19
Ancey C, Gervasoni C, Meunier M (2004) Computing extreme avalanches. Cold Reg Sci Technol 39(2):161–180
Banerjee S, Gelfand A, Carlin B (2003) Hierarchical modeling and analysis for spatial data. Crc Press, Boca Raton
Barbolini M, Keylock C (2002) A new method for avalanche hazard mapping using a combination of statistical and deterministic models. Nat Hazards Earth System Sci 2(3/4):239–245
Barbolini M, Gruber U, Keylock C, Naaim M, Savi F (2000) Application of statistical and hydraulic-continuum dense-snow avalanche models to five real European sites. Cold Regions Sci Technol 31(2):133–149
Bartelt P, Stockli V (2001) The influence of tree and branch fracture, overturning and debris entrainment on snow avalanche flow. Ann Glaciol 32(1):209–216
Beniston M, Diaz H, Bradley R (1997) Climatic change at high elevation sites: an overview. Climatic Change 36(3–4):233–251
Best N, Richardson S, Thomson A (2005) A comparison of Bayesian spatial models for disease mapping. Statistical Methods Med Res 14(1):35–59
Blanchet J, Davison A (2011) Spatial modeling of extreme snow depth. Ann Appl Statistics 5(3):1699–1725
Brooks S, Gelman A (1998) Some issues for monitoring convergence of iterative simulations. In Proceedings of the section on statistical computing, ASA
Carlin B, Louis T (1997) Bayes and empirical Bayes methods for data analysis. Statistics Comput 7(2):153–154
Chopin N (2011) Fast simulation of truncated Gaussian distributions. Statistics Comput 21(2):275–288
Cope EW (2011) Penalized likelihood estimators for truncated data. J Statistical Plan Inference 141(1):345–358
Corona C, Guiot J, Edouard J, Chalié F, Büntgen U, Nola P, Urbinati C (2010) Millennium-long summer temperature variations in the European Alps as reconstructed from tree rings. Climate Past 6(3):379–400
Cressie N, Wikle C (2011) Statistics for spatio-temporal data. Wiley, New York
Diggle P, Ribeiro PJ (2007) Model-based geostatistics. Springer Science+Business Media, Berlin
Durand Y, Laternser M, Giraud G, Etchevers P, Lesaffre B, Mérindol L (2009) Reanalysis of 44 Yr of climate in the French Alps (1958–2002): methodology, model validation, climatology, and trends for air temperature and precipitation. J Appl Meteorol Climatol 48(3):429–449
Eckert N, Parent E, Belanger L, Garcia S (2007) Hierarchical Bayesian modelling for spatial analysis of the number of avalanche occurrences at the scale of the township. Cold Regions Sci Technol 50(1):97–112
Eckert N, Parent E, Naaim M, Richard D (2008) Bayesian stochastic modelling for avalanche predetermination: from a general system framework to return period computations. Stoch Environl Res Risk Assess 22(2):185–206
Eckert N, Parent E, Faug T, Naaim M (2009) Bayesian optimal design of an avalanche dam using a multivariate numerical avalanche model. Stoch Environ Res Risk Assess 23(8):1123–1141
Eckert N, Parent E, Kies R, Baya H (2010) A spatio-temporal modelling framework for assessing the fluctuations of avalanche occurrence resulting from climate change: application to 60 years of data in the northern French Alps. Climatic Change 101(3–4):515–553
Eckert N, Keylock C, Castebrunet H, Lavigne A, Naaim M (2013) Temporal trends in avalanche activity in the French Alps and subregions: from occurrences and runout altitudes to unsteady return periods. J Glaciol 59(213):93–114
Fuentes M (2006) Testing for separability of spatial-temporal covariance functions. J Statistical Plan Inference 136(2):447–466
Garcia-Papani F, Uribe-Opazo M, Leiva V, Aykroyd R (2016) Birnbaum–Saunders spatial modelling and diagnostics applied to agricultural engineering data. Stoch Environ Res Risk Assess:1–20
Gaume J, Eckert N, Chambon G, Naaim M, Bel L (2013) Mapping extreme snowfalls in the French Alps using max-stable processes. Water Resour Res 49(7):1079–1098
Gelman A (2005) Analysis of variance–why it is more important than ever. Ann Statistics 33(1):1–53
Gelman A, Carlin J, Stern H, Rubin D (2004) Bayesian data analysis. Chapman & Hall/CRC, London
Gilks W, Wild P (1992) Adaptive rejection sampling for Gibbs sampling. Appl Statistics 41(2):337–348
Gilks W, Best N, Tan K (1995a) Adaptive rejection metropolis sampling within Gibbs sampling. Applied Statistics:455–472
Gilks W, Richardson S (1995b) Markov Chain Monte Carlo in Practice. Chapman & Hall/CRC Interdisciplinary Statistics, Taylor & Francis, London
Griffiths W (2004) A Gibbs’ sampler for the parameters of a truncated multivariate normal distribution. In: Becker R, Hurn S (eds) Contemporary issues in economics and econometrics: theory and application. Edward Elgar Pub, Camberley, pp 75–91
Jamieson B, Margreth S, Jones A (2008) Application and limitations of dynamic models for snow avalanche hazard mapping. In: Proceedings of the ISSW, pp 730–739
Keylock C (2005) An alternative form for the statistical distribution of extreme avalanche runout distances. Cold Regions Sci Technol 42(3):185–193
Lavigne A, Bel L, Parent E, Eckert N (2012) A model for spatio-temporal clustering using multinomial probit regression: application to avalanche counts. Environmetrics 23:522–534
Lavigne A, Eckert N, Bel L, Parent E (2015) Adding expert contributions to the spatiotemporal modelling of avalanche activity under different climatic influences. J Royal Statistical Soc: series C (Applied Statistics). doi:10.1111/rssc.12095
Leiva V, Ferreira M, Gomes M, Lillo C (2015) Extreme value Birnbaum-Saunders regression models applied to environmental data. Stoch Environ Res Risk Assess 30:1–14
Lied K, Bakkehøi S (1980) Empirical calculations of snow-avalanche run-out distance based on topographic parameters. J Glaciol 26:165–177
Matérn B (1960) Spatial variation, vol 49. Meddelanden från Statens Skogsforskningsinstitut [2nd Edition (2013), vol 36], Springer Science & Business Media, Berlin
McClung D (2001) Extreme avalanche runout: a comparison of empirical models. Canadian Geotech J 38(6):1254–1265
McClung D, Lied K (1987) Statistical and geometrical definition of snow avalanche runout. Cold Regions Sci Technol 13(2):107–119
Meunier M, Ancey C (2004) Towards a conceptual approach to predetermining long-return-period avalanche run-out distances. J Glaciol 50(169):268–278
Munoz F, Pennino MG, Conesa D, López-Quílez A, Bellido JM (2013) Estimation and prediction of the spatial occurrence of fish species using Bayesian latent Gaussian models. Stoch Environ Res Risk Assess 27(5):1171–1180
Naaim M, Faug T, Naaim F, Eckert N (2010) Return period calculation and passive structure design at the Taconnaz avalanche path. France. Ann Glaciol 51(54):89–97
Naaim M, Durand Y, Eckert N, Chambon G (2013) Dense avalanche friction coefficients: influence of physical properties of snow. J Glaciol 59:771–782
Pérez-Cruz F (2008) Kullback-Leibler divergence estimation of continuous distributions. In: IEEE international symposium on information theory, ISIT 2008, IEEE, pp 1666–1670
Rabatel A, Letréguilly A, Dedieu J, Eckert N (2013) Changes in glacier equilibrium-line altitude (ELA) in the western Alps over the 1984–2010 period: evaluation by remote sensing and modeling of the morpho-topographic and climate controls. Cryosphere 7:1455–1471
Rodriguez-Yam G, Davis R, Scharf L (2004) Efficient Gibbs sampling of truncated multivariate normal with application to constrained linear regression. Unpublished Manuscript
Sigurôsson S, Jónasson K, Arnalds P (1998) Transferring avalanches between paths. In: Hestnes E (ed) Proceedings of the anniversary conference 25 years of snow avalanche research, Voss, pp 12–16 May, Norvegian Geotechnical Institute, publication 203, pp 259–263
Smith M, McClung D (1997) Characteristics and prediction of high-frequency avalanche runout. Arctic Alpine Res 29(3):352–357
Speckman P, Sun D (2001) Bayesian nonparametric regression and autoregression priors
Stein ML (1992) Prediction and inference for truncated spatial data. J Comput Graphical Statistics 1(1):91–110
Vanem E, Huseby A, Natvig B (2012) A Bayesian hierarchical spatio-temporal model for significant wave height in the North Atlantic. Stoch Environ Res Risk Assess 26(5):609–632
Wahba G (1978) Improper priors, spline smoothing and the problem of guarding against model errors in regression. J Royal Statistical Soc Series B (Methodological) 40(3):364–372
Zalina M, Desa M, Nguyen V, Kassim A (2002) Selecting a probability distribution for extreme rainfall series in Malaysia. Water Sci & Technol 45(2):63–68
Zhou X, Giacometti R, Fabozzi FJ, Tucker AH (2014) Bayesian estimation of truncated data with applications to operational risk measurement. Quantitative Finance 14(5):862–888
Acknowledgments
The authors thank the ANR research program MOPERA (Modlisation probabiliste pour l’ Etude du Risque d’Avalanche http://www.avalanches.fr/mopera-projet/) for funding this work.
Author information
Authors and Affiliations
Corresponding author
Appendices
Appendix 1: Empirical study of the spatio-temporal covariance structure
In order to investigate the presence of spatio-temporal dependence in runout altitudes, empirical covariance matrices are estimated in the non-separable, separable and additive cases (Fig. 9). Let \(Y_{ict}\) be the runout altitude of avalanche i on path c the year t. The empirical spatio-temporal covariance at spatial lag \([h_{k-1},h_k]\) and time lag \([\tau _{l-1},\tau _l]\) is given by:
with \(N(h_k,\tau _l)\) the set of avalanche pairs \((ict,i'c't')\) such that the distance between sites c and \(c'\) is in the interval \([h_{k-1},h_k]\) and the lag time \(t-t'\) belongs to the interval \([\tau _{l-1},\tau _l]\). \(|N(h_k,\tau _l)|\) is the number of pairs. The reference mean \(\hat{\mu }_Y(c)\) is calculated by linear regression on the valley altitude, the best covariate: \(\hat{\mu }_Y(c)=6.3+ 7.4 10^{-4}h_c\). From the empirical non-separable covariance matrix, the additive and separable covariance matrices have been estimated by least squares. In details, for each spatial and time lag, estimates of \(\gamma ^{T}_l\) and \(\gamma ^S_k\) are found such that:
for the additive covariance matrix, and:
for the separable covariance matrix, where \(\epsilon _{kl}\) is the error.
Figure 9 shows that the non-separable covariance matrix is similar to the separable and additive ones, essentially because the spatial dependence is much higher than the temporal dependence. This pleads further in favor of our choice of a simple model with additive effects.
Appendix 2: Decomposition of the predictive distribution
First we integrate the annual predictive distribution \([\mathbf {Y}_{c_0t}|\mathbf {Y}]\) for time t and paths \(c_0\) over the time to make prediction for the entire study period:
By using the hierarchical structure of the model we can express this distribution conditionally to the process level:
Let \(\varvec{\theta }_1=(B_t,\alpha ,\sigma _{c_0}^2)'\), we decompose the distribution \([\mathbf {C}_{c_0},\varvec{\theta }_1|\mathbf {Y}]\) in the same way to make the spatial term \(\mathbf {A}_{c_0}\) appear,
Finally, we express \([\mathbf {A}_{c_0},\varvec{\theta }_2|\mathbf {Y}]\), with \(\varvec{\theta }_2=(a,b,o_S,\rho ^2)'\), conditionally to \(\mathbf {A}\) rather than \(\mathbf {Y}\):
The joint Gaussian distribution for the vector \([\mathbf {A},\mathbf {A}_{c_0}|\phi ,\tau ^2]\) is:
where \(\varvec{\Sigma }_{\phi }^{..}\) and \(\varvec{\Sigma }_{\phi }^{00}\) are the variance-covariance matrices of \(\mathbf {A}\) and \(\mathbf {A}_{c_0}\), respectively, and where \(\varvec{\Sigma }_{\phi }^{.0}\) and \(\varvec{\Sigma }_{\phi }^{0.}\) are covariance matrices between elements of \(\mathbf {A}\) and \(\mathbf {A}_{c_0}\). Then the formulas for conditional Gaussian distributions give:
By combining Eqs. (11) and (12) within Eq. (10), and using the Bayes’ rule we obtain
with \(\varvec{\theta }_3\) the vector \((\mathbf {A},\phi ,\tau ^2)'\).
Appendix 3: Algorithm details
We detail the algorithm to draw \(\mathbf {C}\) under the constraint \(\sum _{c=1}^{n_c}C_c=0\), the same method is used to draw \(\mathbf {B}\) under the constraint \(\sum _{t=1}^{n_t}B_t=0\).
Let \(\mathbf {X}_C\) the \(n\times n_c\) matrix such that \({\mathbf {X}_C}_{ij}=1\) if the avalanche i occurs on path j, 0 either, \(\mathbf {D}\) the diagonal variance-covariance matrix of \(\mathbf {Z}\), \(\mathbf {D}=\text {diag}(\sigma _{c_i}^2)_{i=1,n}\), with \(c_i\) the path label for avalanche i, and \(\mathbf {1}_c\) the \(n_c\) length column vector of ones.
The complete posterior distribution of \(\mathbf {C}\) is given by,
At a second step we write the joint distribution of \(\mathbf {C}|\mathbf {Z},\mathbf {B},\alpha ,\mathbf {D},E[\mathbf {C}],\rho ^2\), noted \(\mathbf {C}|\varvec{\gamma }\) and its constraint \(\mathbf {1}_{c}\mathbf {C}=0\),
The conditional distribution is then:
Only the \(n_c-1\) first components of \(\mathbf {C}\) are sampled, \(C_{n_c}\) is given by \(C_{n_c}=-\sum _{c=1}^{n_c-1}C_c\).
Then we constrain the mean runout altitude by path, \(\alpha +C_c\). We have \(h_c<\alpha +C_c<s_c\) for all \(c\in \{1,\cdots n_c\}\). In a Gibbs sampler, \(\mathbf {C}\) and \(\alpha\) are drawn successively in their complete conditional distributions. We describe the method to take into account the constraint for \(\mathbf {C}\), \(\alpha\) is sampled in a similar way. The \(n_c-1\) first components of \(\mathbf {C}\) must fullfill the constraints:
that we can write matrix-wise \(\mathbf {V}\mathbf {C}<\mathbf {v}\) with \(\mathbf {V}\) the \(k\times n_{c}-1\) matrix of constraints. Here \(k=2n_c\), since there are two constraints for each line of the system Eq. (15), one bears on \(s_c\), the other one on \(h_c\). The posterior distribution of \(\mathbf {C}\) is thus a multivariate truncated distribution whose mean and variance for the complete version are given by Eq. (14) and whose support \(S=\{\mathbf {x}\in \mathbb {R}^{n_c-1} {:}\, \mathbf {V}\mathbf {x}<\mathbf {v}\}\). We follow the two-step algorithm of Rodriguez-Yam et al. (2004). In the first step, we remove the dependence between the components of \(\mathbf {C}\) by considering the vector \(\mathbf {w}=\mathbf {L}\mathbf {C}\), with \(\mathbf {L}\) such that \(\mathbf {L}\varvec{\Sigma }_0 \mathbf {L}'=\mathbf {I}_{n_c-1}\) and \(\mathbf {I}_{n_c-1}\) the identity matrix. It has the following truncated distribution:
In the second step, each component of \(\mathbf {w}\) is drawn successively conditionally to the others in a Gibbs sampling. In order to avoid the computation of the entire vector of constraints \(\mathbf {V}\mathbf {L}^{-1} \mathbf {w}\) at each draw, we may notice that for each constraint \(i \in \{1, \ldots k\}\), and each component of \(\mathbf {w}\) \(j_0\in \{1, \ldots n_c-1\}\)
The constraint i for the \(j_0\) component of \(\mathbf {w}\) is written \(u_{i}^{j_0}w_{j_0} < v_i - u_{i}^{-j_0}\). To update \(w_{j_0}\), one only has to compute \(u_{i}^{j_0}\) for each constraint. This step demands \(k(n_c-1)\) elementary operations, instead of \(k(n_c-1)^2\) operations when \(\mathbf {V}\mathbf {L}^{-1}\) is computed naively.
Finally we implement the following algorithm for sampling the \(n_c-1\) components of \(\mathbf {C}\) from their truncated Gaussian distributions:
-
Compute the vector \(\mathbf {u}= \mathbf {V}\mathbf {C}\);
-
Initiate the matrix \(\mathbf {L}\), its inverse \(\mathbf {L}^{-1}\), and the vector \(\mathbf {w}=\mathbf {L}\mathbf {C}\);
-
For each j in \(\{1,\ldots n_c-1\}\)
-
Compute the k-dimension vectors \(\mathbf {u}^j=(u_1^j,\ldots u_k^j)\), and \(\mathbf {u}^{-j}= \mathbf {u}- u^jw_j\);
-
Find the interval \([ a_j, b_j ]\) in which \(w_j\) is drawn to satisfy the k constraints;
-
Sample \(w_j\) in the normal truncated distribution \(\mathcal {N}((\mathbf {L}\mathbf {m}_{0})_j,1)\) with support \([ a_j, b_j ]\). Here, we use the method and the code proposed by Chopin (2011);
-
Update \(\mathbf {u}\) as \(\mathbf {u}=\mathbf {u}^{-j}+\mathbf {u}^jw_j\);
-
-
Set \(\mathbf {C}\) as \(\mathbf {C}=\mathbf {L}^{-1}\mathbf {w}\).
Rights and permissions
About this article
Cite this article
Lavigne, A., Eckert, N., Bel, L. et al. Modelling the spatio-temporal repartition of right-truncated data: an application to avalanche runout altitudes in Hautes-Savoie. Stoch Environ Res Risk Assess 31, 629–644 (2017). https://doi.org/10.1007/s00477-016-1301-z
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00477-016-1301-z