Sensitivity analysis of a sensitivity analysis: We are likely overlooking the impact of distributional assumptions

A mechanistic approach to include climate change and unplanned urban sprawl in landslide susceptibility maps

Empirical evidence shows that climate, deforestation and informal housing (i.e. unregulated construction practices typical of fast-growing developing countries) can increase landslide occurrence. However, these environmental changes have not been considered jointly and in a dynamic way in regional or national landslide susceptibility assessments. This gap might be due to a lack of models that can represent large areas (>100km²) in a computationally efficient way, while simultaneously considering the effect of rainfall infiltration, vegetation and housing. We therefore suggest a new method that uses a hillslope-scale mechanistic model to generate regional susceptibility maps under changing climate and informal urbanisation, which also accounts for existing uncertainties. An application in the Caribbean shows that the landslide susceptibility estimated with the new method and associated with a past rainfall-intensive hurricane identifies ~67.5 % of the landslides observed after that event. We subsequently demonstrate that the hypothetical expansion of informal housing (including deforestation) increases landslide susceptibility more (+20 %) than intensified rainstorms due to climate change (+6 %). However, their combined effect leads to a much greater landslide occurrence (up to +40 %) than if the two drivers were considered independently. Results demonstrate the importance of including both land cover and climate change in landslide susceptibility assessments. Furthermore, by modelling mechanistically the overlooked dynamics between urban growth and climate change, our methodology can provide quantitative information of the main landslide drivers (e.g. quantifying the relative impact of deforestation vs informal urbanisation) and locations where these drivers are or might become most detrimental for slope stability. Such information is often missing in data-scarce developing countries but is key for supporting national long-term environmental planning, for targeting financial efforts, as well as for fostering national or international investments for landslide mitigation.

A trait-based model ensemble approach to design rice plant types for future climate

Crop models are powerful tools to support breeding because of their capability to explore genotype × environment×management interactions that can help design promising plant types under climate change. However, relationships between plant traits and model parameters are often model specific and not necessarily direct, depending on how models formulate plant morphological and physiological features. This hinders model application in plant breeding. We developed a novel trait-based multi-model ensemble approach to improve the design of rice plant types for future climate projections. We conducted multi-model simulations targeting enhanced productivity, and aggregated results into model-ensemble sets of phenotypic traits as defined by breeders rather than by model parameters. This allowed to overcome the limitations due to ambiguities in trait-parameter mapping from single modelling approaches. Breeders' knowledge and perspective were integrated to provide clear mapping from designed plant types to breeding traits. Nine crop models from the AgMIP-Rice Project and sensitivity analysis techniques were used to explore trait responses under different climate and management scenarios at four sites. The method demonstrated the potential of yield improvement that ranged from 15.8% to 41.5% compared to the current cultivars under mid-century climate projections. These results highlight the primary role of phenological traits to improve crop adaptation to climate change, as well as traits involved with canopy development and structure. The variability of plant types derived with different models supported model ensembles to handle related uncertainty. Nevertheless, the models agreed in capturing the effect of the heterogeneity in climate conditions across sites on key traits, highlighting the need for context-specific breeding programmes to improve crop adaptation to climate change. Although further improvement is needed for crop models to fully support breeding programmes, a trait-based ensemble approach represents a major step towards the integration of crop modelling and breeding to address climate change challenges and develop adaptation options.

Global Sensitivity Analysis with Mixtures: A Generalized Functional ANOVA Approach

This work investigates aspects of the global sensitivity analysis of computer codes when alternative plausible distributions for the model inputs are available to the analyst. Analysts may decide to explore results under each distribution or to aggregate the distributions, assigning, for instance, a mixture. In the first case, we lose uniqueness of the sensitivity measures, and in the second case, we lose independence even if the model inputs are independent under each of the assigned distributions. Removing the unique distribution assumption impacts the mathematical properties at the basis of variance-based sensitivity analysis and has consequences on result interpretation as well. We analyze in detail the technical aspects. From this investigation, we derive corresponding recommendations for the risk analyst. We show that an approach based on the generalized functional ANOVA expansion remains theoretically grounded in the presence of a mixture distribution. Numerically, we base the construction of the generalized function ANOVA effects on the diffeomorphic modulation under observable response preserving homotopy regression. Our application addresses the calculation of variance-based sensitivity measures for the well-known Nordhaus' DICE model, when its inputs are assigned a mixture distribution. A discussion of implications for the risk analyst and future research perspectives closes the work.

Sensitivity analysis methods in the biomedical sciences

Sensitivity analysis is an important part of a mathematical modeller's toolbox for model analysis. In this review paper, we describe the most frequently used sensitivity techniques, discussing their advantages and limitations, before applying each method to a simple model. Also included is a summary of current software packages, as well as a modeller's guide for carrying out sensitivity analyses. Finally, we apply the popular Morris and Sobol methods to two models with biomedical applications, with the intention of providing a deeper understanding behind both the principles of these methods and the presentation of their results.

Tailoring parameter distributions to specific germplasm: impact on crop model-based ideotyping

Crop models are increasingly used to identify promising ideotypes for given environmental and management conditions. However, uncertainty must be properly managed to maximize the in vivo realizability of ideotypes. We focused on the impact of adopting germplasm-specific distributions while exploring potential combinations of traits. A field experiment was conducted on 43 Italian rice varieties representative of the Italian rice germplasm, where the following traits were measured: light extinction coefficient, radiation use efficiency, specific leaf area at emergence and tillering. Data were used to derive germplasm-specific distributions, which were used to re-run a previous modelling experiment aimed at identifying optimal combinations of plant trait values. The analysis, performed using the rice model WARM and sensitivity analysis techniques, was conducted under current conditions and climate change scenarios. Results revealed that the adoption of germplasm-specific distributions may markedly affect ideotyping, especially for the identification of most promising traits. A re-ranking of some of the most relevant parameters was observed (radiation use efficiency shifted from 4th to 1st), without clear relationships between changes in rankings and differences in distributions for single traits. Ideotype profiles (i.e., values of the ideotype traits) were instead more consistent, although differences in trait values were found.

Sensitivity analysis of the Aquacrop and SAFYE crop models for the assessment of water limited winter wheat yield in regional scale applications

Process-based models can be usefully employed for the assessment of field and regional-scale impact of drought on crop yields. However, in many instances, especially when they are used at the regional scale, it is necessary to identify the parameters and input variables that most influence the outputs and to assess how their influence varies when climatic and environmental conditions change. In this work, two different crop models, able to represent yield response to water, Aquacrop and SAFYE, were compared, with the aim to quantify their complexity and plasticity through Global Sensitivity Analysis (GSA), using Morris and EFAST (Extended Fourier Amplitude Sensitivity Test) techniques, for moderate to strong water limited climate scenarios. Although the rankings of the sensitivity indices was influenced by the scenarios used, the correlation among the rankings, higher for SAFYE than for Aquacrop, assessed by the top-down correlation coefficient (TDCC), revealed clear patterns. Parameters and input variables related to phenology and to water stress physiological processes were found to be the most influential for Aquacrop. For SAFYE, it was found that the water stress could be inferred indirectly from the processes regulating leaf growth, described in the original SAFY model. SAFYE has a lower complexity and plasticity than Aquacrop, making it more suitable to less data demanding regional scale applications, in case the only objective is the assessment of crop yield and no detailed information is sought on the mechanisms of the stress factors affecting its limitations.