Machine-learning methods to assess the effects of a non-linear damage spectrum taking into account soil moisture on winter wheat yields in Germany

Peichl, Michael; Thober, Stephan; Samaniego, Luis; Hansjürgens, Bernd; Marx, Andréas

doi:10.5194/hess-25-6523-2021

Cited by 18 publications

(10 citation statements)

References 78 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Too wet conditions in March are found to be an impact-relevant factor, in agreement with Peichl et al (2021). SMI-Total adds complementary information to monthly SPEI.…”

Section: Discussionsupporting

confidence: 74%

“…In addition to the monthly SPEI, a metric of growing season drought magnitude was computed as the sum of SPEI-1 < -0.5 over the period between March and July (SPEI-Magnitude). Regarding soil moisture droughts, the model-based German drought monitor developed at the Helmholtz Centre for Environmental Research (UFZ) is the most established regional product (Samaniego et al, 2013;Zink et al, 2016;Boeing et al, 2022) and has already been used for similar purpose (Peichl et al, 2021). We use the monthly soil moisture index (SMI) and aggregated soil drought magnitude (SMI-Magnitude) in the top soil (25 cm), again from March to July.…”

Section: Hazard Indicators: Spei and Smimentioning

confidence: 99%

See 1 more Smart Citation

Exploring drought hazard, vulnerability, and related impacts to agriculture in Brandenburg

Brill,

Alencar,

Zhang

et al. 2024

Preprint

View full text Add to dashboard Cite

Abstract. Adaptation to an increasingly dry regional climate requires spatially explicit information about current and future risks. Existing drought risk studies often rely on expert-weighted composite indicators, while empirical evidence on impact-relevant factors is still scarce. The aim of this study is to investigate to what extent hazard and vulnerability indicators can explain observed agricultural drought impacts via data-driven methods. We focus on the German federal state of Brandenburg, 2013–2022, including several consecutive drought years. As impact indicators we use thermal-spectral anomalies (LST/NDVI) on field level, and empirical yield gaps from reported statistics on county level. Empirical associations to the impact indicators on both spatial levels are compared. Non-linear models explain up to about 60 % variance in the yield gap data, with lumped models for all crops being more stable than models for individual crops, and models for the drought years performing better than for the pre-drought years. Meteorological drought in June and soil quality are selected as strongest impact-relevant factors. Rye is found less vulnerable than wheat, despite growing on poorer soils. LST/NDVI only weakly relates to our empirical yield gaps. We recommend comparing different impact indicators on multiple scales to proceed with the development of empirically grounded risk maps.

show abstract

“…Too wet conditions in March are found to be an impact-relevant factor, in agreement with Peichl et al (2021). SMI-Total adds complementary information to monthly SPEI.…”

Section: Discussionsupporting

confidence: 74%

Section: Hazard Indicators: Spei and Smimentioning

confidence: 99%

Exploring drought hazard, vulnerability, and related impacts to agriculture in Brandenburg

Brill,

Alencar,

Zhang

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…The reason for selecting upper 30 cm is from the potential impact point of view: flash droughts that affect the agricultural areas with crops their dominant root biomass within the depth of 30 cm (crops: ≈ 70% of the roots in the top 30 cm; Temperate grassland: ≈ 83% of the roots in the top 30 cm) (Jackson et al 1996) prominently. Additionally, it was shown by Peichl et al (2021) across Germany that the top 25 cm of soil moisture is a better yield predictor than the total soil column.…”

Section: Methodsmentioning

confidence: 99%

On the role of antecedent meteorological conditions on flash drought initialization in Europe

Shah

Kumar

Samaniego

et al. 2023

Environ. Res. Lett.

Self Cite

View full text Add to dashboard Cite

The fast depletion of soil moisture in the top soil layers characterizes flash drought events. Due to their rapid onset and intensification, flash droughts severely impact ecosystem productivity. Thus understanding their initialization mechanisms is essential for improving the skill of drought forecasting systems. Here, we examine the role of antecedent meteorological conditions that lead to flash droughts across Europe over the last 70 years (1950--2020) using ERA5 dataset. We find two major flash-drought types based on a sequence of development of antecedent hydro-meteorological conditions. The first type is characterized by a joint occurrence of two mechanisms, a decline of precipitation in conjunction with an increase of the evaporative demand, both occurring before the onset of a flash drought event. The second type, on the contrary, is characterized by high precipitation preceding the event's start, followed by a sudden precipitation deficit combined with an increase in evaporative demand at the onset of the drought. Both drought types showed increased occurrence and higher spatial coverage over the last 70 years; the second drought type has increased at a much faster rate compared to the first one specifically, over Central Europe and the Mediterranean region. Overall our study highlights the differences between the two types of flash droughts, related to varying antecedent meteorological conditions, and their changes under recent climate warming.

show abstract

“…The line graph in the ALE plots described how the prediction of RFR model changes over small intervals of each factor (Apley & Zhu, 2020). The local change effects are accumulated across all intervals and the mean effect is centered at zero (Macdonald et al., 2022; Peichl et al., 2021). The ALE plots are implemented based “ALEPlot” package on R studio (https://cran.r-project.org/web/packages/ALEPlot/index.html).…”

Section: Methodsmentioning

confidence: 99%

Flood Defense Standard Estimation Using Machine Learning and Its Representation in Large‐Scale Flood Hazard Modeling

Zhao

Bates

Neal

et al. 2023

Water Resources Research

View full text Add to dashboard Cite

Floods are the most frequent type of natural hazards worldwide, and have caused significant loss of life and severe economic impacts on populations and property during the past two decades (CRED and UNDRR, 2020). To reduce the negative impact of floods, numerous types of flood defenses, such as levee systems, have been built to protect cities, towns, and farms in almost every country (Ubilla et al., 2008;Z. Wang & Liu, 2019). According to a report by the United States (U.S.) Army Corps of Engineers (USACE), approximately 11 million people and $1.3 trillion of property value existed in flood-defended areas in the U.S. as of 2018 (USACE, 2018). The number and standard of flood defenses continuously improve over time to meet societal needs and keep pace with rapid urbanization in floodplains (Leonard, 2008;T. Zhu et al., 2007). Flood defenses have significantly changed the regional flooding distribution and also residents' exposure (Di Baldassarre et al., 2009;Ludy & Kondolf, 2012), and this needs to be considered in flood hazard assessment.With the increase of computing power and advances in remote sensing techniques, it is now possible to map flood hazards on a large scale at high resolution (<100 m) (Ward et al., 2015). High resolution global hydrography datasets, such as MERIT Hydro (Yamazaki et al., 2019) and HydroBASINS (Lehner & Grill, 2013) have been released; however, information on detailed flood defenses for most rivers in the world is severely limited (Aerts et al., 2020;Sampson et al., 2015). Existing state-of-the-art global flood hazard models either assume a simplified high flood defense standard (FDS) or assume no protection when applied (Aerts et al., 2020;Scussolini et al., 2016;Ward et al., 2015). This assumption causes the misestimations of flood hazards when the flooded areas are actually protected by existing flood defenses, and therefore induces a distorted flood hazard and risk assessment. The first global flood defense database (called FLOPROS) was built by collecting FDS data worldwide at the sub-country scale (Scussolini et al., 2016). FLOPROS assumes that the FDSs in a vast area are the same (i.e., most states in the US or all of Australia have the same FDS) and ignores the heterogeneity of FDSs between rivers. The coarse resolution of the FDS data in FLOPROS cannot meet the requirement of large-scale flood hazard modeling at high resolution.

show abstract

Machine-learning methods to assess the effects of a non-linear damage spectrum taking into account soil moisture on winter wheat yields in Germany

Cited by 18 publications

References 78 publications

Exploring drought hazard, vulnerability, and related impacts to agriculture in Brandenburg

Exploring drought hazard, vulnerability, and related impacts to agriculture in Brandenburg

On the role of antecedent meteorological conditions on flash drought initialization in Europe

Flood Defense Standard Estimation Using Machine Learning and Its Representation in Large‐Scale Flood Hazard Modeling

Contact Info

Product

Resources

About