5th iLEAPS Science Conference Abstracts - B2

Abstracts – Session B2

Changing water cycle in the food baskets of the world (joint session with GEWEX)

B201ORAL-0174: Including Management of the Water Cycle in Land Surface Models

Richard Harding1, Jan Polcher2

1CEH, Wallingford, United Kingdom 2LMD, Paris, France

Most of our major rivers are moderately or severely impacted by man’s activities – leading to reduction of flows and changes in the seasonal regime caused by the dams. Irrigation use is the major water extraction from rivers, reservoirs and groundwater. The irrigated area of the world occupies only a few percent of the land area but produces about 40% of our food and regionally can be large, particularly in semiarid ‘hotspots’. The area and intensity of irrigation is likely to increase in the future as we struggle to feed up to 11 billion people in a few decades.
There are many tens of thousands of dams worldwide which store approximately 8000 km3 of water – a significant fraction of the global annual river flow. Some of these impacts are directly coupled to the atmosphere, influencing, for example, temperature extremes and evaporation and, ultimately, cloud and rainfall.
A GEWEX workshop, held over three days in September 2016, brought together over 40 scientists from a wide range of disciplines, land surface scientists, hydrologists, meteorologists and agronomists, to review the current state of models and supporting data, and to sketch out a path for including man’s management of water in future land surface models.
The workshop identified the need to include irrigation, reservoirs and water diversions within operational versions of Land Surface Models. Groundwater stores and recharge should also be an aspiration. Further sophistication might include different crop types and irrigation practices. For future prediction some estimate must be made of future extent and practice or irrigation – this could be taken from scenarios or Integrated Assessment Models but ultimately the growth of the extent of irrigation will be limited by the availability of water. Thus a full earth system model should include population, food production, infrastructure and the environmental limits.

B202ORAL-0170: Global assessment of water policy challenges under uncertainty in water scarcity projections

Peter Greve1, Taher Kahil1, Junko Mochizuki1, Thomas Schinko1, Yoshihide Wada1

1International Institute for Applied Systems Analysis, Laxenburg, Austria

Water scarcity is a critical environmental issue worldwide, which has been primarily driven by a significant increase in water extractions (especially from irrigation) during the last century. In the coming decades, climate and societal changes are projected to further exacerbate water scarcity conditions in many regions around the world. At present, one important question for the pending policy debate is the identification of water policy interventions that could address upcoming water scarcity problems in the presence of substantial uncertainties. Hence, anticipating the long term environmental conditions and economic and agricultural needs in the context of associated uncertainties is highly relevant for an appropriate decision-making process. Here we assess in a probabilistic approach global water scarcity projections following different socioeconomic pathways (SSPs) and climate-forcing scenarios (RCPs) within the first half of the 21st century. By utilizing a comprehensive set of global water scarcity projections we identify, besides trends in average water scarcity, changes in the uncertainty range of anticipated water scarcity conditions and further attribute the underlying sources of uncertainty. Our results show that both average water scarcity and the associated range of uncertainty are generally increasing in most regions of the world, including many intensively cultivated agricultural areas. Most of the uncertainty in the projections can be attributed to differences in the employed hydrological impact models rather than to projected socioeconomic development. Based on these results, we develop a general policy and decision-making framework that provides a valuable contribution to effective policymaking by identifying four representative clusters of specific water policy challenges and needs. Based on this framework we further specifically outline how uncertain projections of water scarcity could be addressed within the food production sector.

B203 ORAL-0230: Evaluation of earth2observe global Water Resources Re-analysis using global EO datasets and flux tower evapotranspiration data to focus on dry-down processes

Alberto Martinez de la Torre1, Eleanor Blyth1

1CEH, Wallingford, United Kingdom

EartH2Observe is an European project that brings together the modelling (LSMs and global hydrological models) and EO communities, integrating available global earth observations, in-situ datasets and models and building a global Water Resources Re-analysis (WRR) dataset of significant length (1979-2015). Here we evaluate this WRR, focusing in the current model capacity to represent the water cycle and drying processes over food producing regions.
For the evaluation we use the ILAMB global benchmarking system (that analyses model performance against a series of global hydrological EO datasets) and a new dry-down metric: a physical parameter that characterizes model dry-down processes in water limited conditions, in terms of evapotranspiration decay during dry events (10+ day periods of no precipitation). For the dry-down evaluation, we use a set of flux tower sites that provides half-hourly evapotranspiration data and represents different land covers and climates around the world.
Our dry-down analysis characterizes different model responses to water limited conditions and provides a potential constraint for future climate change model projections, linking global model processes to local behaviour through a physical parameter.

B204 - RAL-0132: Estimating fresh water inputs into the Mediterranean sea under global change

Fuxing WANG1, Jan Polcher1

1Laboratoire de Météorologie Dynamique, Ecole Polytechnique, Palaiseau, France

Fresh water is expected to become an increasingly scarce resource in the future, which is the expected consequence of the combination of climate change and increasing water uses. The Mediterranean is expected to be one of the most prominent and vulnerable climate change “hotspots” of the 21st century. Previous studies reported that negative trends of annual stream flow (drying trends) are detected in the southern and eastern Mediterranean regions. The freshwater into Mediterranean sea is important factor in the large scale circulation of this closed basin as well as for maintaining marine and coastal ecosystems. Due to the uncertainties from un-gauged rivers, human activities, and measurement of water flow at river outlet, estimating the freshwater input into the Mediterranean Sea is associated with large uncertainties. Previous studies estimate freshwater inflow into the Mediterranean Sea either by combining simple annual water balance and where available observations (only providing annual mean discharge values) or land surface model (LSM) forced by atmospheric conditions (affected by model uncertainties).
In the proposed study use data assimilation techniques to merge model output (ORCHIDEE LSM developed at Institut Pierre Simon Laplace) and observed river discharge from Global Runoff Data Center (GRDC, about 630 stations) to obtain optimized river discharges for all rivers contributing to the Mediterranean basin. This approach compensates for systematic errors of the model as well as missing processes. It provides estimation of the riverine input into the sea at high temporal and spatial resolution (1979-2013). We will analyze the trend of freshwater inflow into the Mediterranean Sea and river discharge change over different regions (e.g., natural as well as managed basins) with the help of this new data. The new estimates will serve to estimate the impact on the Mediterranean basin of continental water cycle changes and support water balance studies of the region.

B205 - ORAL-0288: Recent and future intensification of the water cycle in the Sahel: implications for agriculture

Christopher Taylor1, 2, Danijel Belusic3, Francoise Guichard4, Douglas Parker5, Theo Vischel6, Olivier Bock7, Phil Harris1, 2, Serge Janicot8, Connie Klein1, Geremy Panthou6, Francois Affholder9, Benjamin Sultan8, Adama Tounkara10

1CEH, Wallingford, United Kingdom 2NCEO, Wallingford, United Kingdom 3SMHI, Norrkoping, Sweden 4CNRM UMR 3589, Météo-France/CNRS, Toulouse, France 5School of Earth and Environment, University of Leeds, Leeds, United Kingdom 6CNRS, Univ. Grenoble Alpes, Institut de Géosciences de l’Environnement (IGE), grenoble, France 7IGN, Universite Paris Diderot, Paris, France 8LOCEAN, Sorbonne University, Paris, France 9CIRAD, Montpellier, France 10ISRA, Dakar, Senegal

In West Africa, any trends in rainfall characteristics have potentially important impacts on agriculture. Crops are predominantly rain-fed, and yields are sensitive to the timing of rainfall within the season, as well as season length and total rainfall. Moreover, changes in rainfall intensity influence runoff and soil erosion, with impacts on yields through nutrient losses.
Here we document observed changes in the intensity of Sahelian Mesoscale Convective Systems (MCS) over the 35 year satellite record. MCS provide around 90% of Sahel rainfall, and can be readily identified from sub-hourly cloud-top temperature images from Meteosat. Defining MCS as large contiguous cold cloud features, we find that the strong interannual and decadal variability in seasonal rainfall is well-captured by the number of MCS at a threshold of -40°C. Superimposed on this however, is a remarkable downward trend in cloud-top temperatures within MCS which has continued since the so-called recovery of seasonal rainfall. When considering a temperature threshold of -70°C, we find a tripling in the frequency of MCS observed over the Sahel during this period. We interpret these features as a tendency towards more intense convection in late afternoon and evening, in turn creating larger, longer-lived systems overnight.
The trend towards more intense MCS is well-correlated with rising global land temperatures, but not Sahelian temperatures. Instead we argue that the rapidly warming Sahara may provide the regional link to global warming, via changes in wind shear and the Saharan Air Layer. The meridional temperature gradient spanning the Sahel has increased in recent decades, consistent with anthropogenic emissions in the CMIP5 ensemble, and is set to continue throughout the 21st century.

This suggests that the Sahel will experience particularly marked increases in extreme rain. We will discuss the challenges that this trend poses for agriculture in the region, and potential mitigation options.

B206ORAL-0323: The sensitivity of future water availability in explicit vs parameterised convection schemes over Africa

Sonja Folwell1, Christopher Taylor1, Rachel Stratton2, Simon Tucker2

1Centre for Ecology & Hydrology, Wallingford, United Kingdom 2Met Office, Exeter, United Kingdom

Climate models are an essential tool in understanding and mitigating against the impacts of climate change on water availability and extremes. Projected changes in these extremes have important implications, for example, on urban planning and the agricultural sectors. One particular difficulty in translating the global warming signal to regional changes in water availability is through the representation of rainfall events. Climate models often underestimate the intensity of rainfall, particularly where convective rainfall is dominant. Recent developments in climate model convection schemes have resulted in large improvements in the distribution of daily and sub-daily rainfall intensities. Under the IMPALA project, the UK Met. Office has undertaken a set of ambitious pan-Africa experiments; a high resolution (4.5 km) convection permitting simulation and coarser scale (25 km) parameterised convection simulation for current and future climates. These two simulations provide contrasting depictions of rainfall frequency and intensity and give an important opportunity to reassess the sensitivity of the surface water balance (soil moisture and runoff) across Africa to a warmer climate, and the consequences for rain-fed agriculture in the region.

B207ORAL-0290: Modelling the changing water balance in West Africa

Peter Anthony Cook1, Emily Black 1, Anne Verhoef2

1NCAS-Climate, University of Reading, Reading, United Kingdom 2University of Reading, Department of Geography and Environmental Science, Reading, United Kingdom

The vulnerability of many people in West Africa to hydrological hazard means that it is critical to understand how the monsoon will change in the future. We are using the JULES land-surface model, driven by data from a Met Office Unified Model (GA3) ensemble of 25km horizontal resolution atmosphere-only model runs for both present (~2000) and future (~2100) climate. Our study examines the changes in the amount of rain and its intensity, in the spatial distribution and timing of the monsoon, and in the water balance under a high emissions scenario. It is projected that rainfall will increase and become more intense, the monsoon will shift northwards and occur later in the year, and evapo-transpiration will be reduced, while runoff and drainage will increase. We are also determining how the water balance depends on rainfall intensity, soil makeup and vegetation coverage. This work is part of the BRAVE2 project, "building understanding of climate variability into planning of groundwater supplies from shallow aquifers in Africa (second phase)", funded under the NERC/DFID/ESRG Unlocking the Potential of Groundwater for the Poor (UPGro) program.

B208ORAL-0031: Reduction of monsoon rainfall in response to past and future land-use and land-cover changes

Benjamin Quesada1, Narayanappa Devaraju2, Nathalie De Noblet-Ducoudré2, Almut Arneth1

1Karlsruhe Institute of Technology, Institute of Meteorology and Climate research, Atmospheric Environmental Research, Garmisch-Partenkirchen, Germany 2Laboratoire des Sciences du Climat et de l'Environnement (LSCE), Gif-sur-Yvette, France

Land-use and land cover changes (LULCC) can have significant biophysical impacts on regional precipitation, including monsoon rainfall. However, most modelling studies are based on the use of only one global or regional climate model, focus on one specific region and/or apply different idealized deforestation/afforestation scenarios (e.g., 100% or 50% of forests replaced by crops or grasslands and vice versa). Here, using global simulations performed with and without LULCC from 5 Earth System Models (ESM), under a common and realistic LULCC scenario (the Representative Concentration Pathway RCP8.5), we find that future LULCC significantly reduce monsoon precipitation in at least 4 (out of 8) monsoon regions.
While monsoon rainfalls are likely to intensify under future global warming, we estimate that biophysical effects of LULCC substantially weaken future projections of monsoons’ rainfall by 9% (Indian region), 12% (East Asian), 32% (South African) and 41% (North African), with an average of ~30% for projections across the global monsoon region. A similar strong contribution is found for biophysical effects of past LULCC to monsoon rainfall changes since the preindustrial period. Rather than remote effects (e.g., ITCZ shifts or meridional heat transport), local land-atmosphere interactions, implying a decrease in evapotranspiration, soil-moisture and clouds along with more anticyclonic conditions, could explain this reduction in monsoon rainfall.
As only ~2/3 of global climate models account for LULCC [IPCC et al., 2013], the current average projections of monsoon rainfall could be overestimated. In consequence, to increase confidence and robustness in monsoon projections and climate mitigation strategies, we stress here the importance of considering carefully LULCC for future projections of the hydrological cycle.

B209 -ORAL-0141: Linking Food to Water from a Biophysical Perspective - A Nexus Approach Applied to South Africa Under Present and Projected Future Climatic Conditions

Roland Schulze1

1University of KwaZulu-Natal, Pietermaritzburg, South Africa

Following brief introductory remarks on the Water-Energy-Food nexus concept, this presentation focuses on the Water-Food Nexus from a South African food and water security perspective, with “food” viewed in its broader context of food, fibre, fruit, fodder, forest and fuel, stressing both feed-forwards and feed-backs between the two and illustrating each case with South African examples, both without and with projected climate change impacts. Issues that are stressed include those related to dryland (rainfed) vs irrigation cropping, climate as a driver for food and water availability, the problem of water quality related to agriculture, the dilemma of different farming systems in South Africa in relation to water, challenges around water and biofuel crops, environmental problems around water and irrigation, water related constraints on crop growth and managed vs.mis-managed farming systems and water security The presentation concludes with a plea to move beyond more conventional tri-partite Water-Energy-Food Nexus thinking by incorporating a “land” component as well as a value chain approach that goes further than the farmgate, the river and the reservoir.

B210ORAL-0313: Towards the representation of human processes in the routing scheme of the ORCHIDEE land surface model

Jan Polcher1, Xudong Zhou2, Patrice Dumas3

1LMD/IPSL/CNRS, Paris, France 2LMD/IPSL/U. Paris Saclay, Paris, France 3CIRED, Paris, France

Land surface models (LSM) used in Earth system models, regional or global, have acquired a capability to simulate the natural water, energy and carbon cycles over the continents. These models simulate the state variables of the surface, the fluxes exchanged with the atmosphere and the river discharge into the oceans. LSMs have known a rapid evolution over the last 20 years and are now the tool of choice to perform re-analysis of surface states and to study land surface processes in interaction with the other components of the Earth system.
These models, in contrast to global hydrological models, mostly neglect the fact that human regulate the water cycle of the continents and the vegetation for the benefit of their agricultural and industrial systems. Especially in the food baskets of the world, the flow of water and vegetation state are highly controlled by humans and have diverged from their natural state. This is not without consequence on the sensitivity of land surfaces to climate variability. These human processes need to be integrated to make LSM predictions for a climate change relevant for society. This evolution will probably be more transformational than when the carbon cycle was introduced.
We will show how this change will be implemented into ORCHIDEE. The introduction of value classes for all surface waterflows allows to implement operating rules for human water usage which maximise the benefits of water for human activities. Simple test cases allow to show this increases the water usage for irrigation and yields more realistic river discharge. The numerical difficulties of implementing human processes in LSMs will also be discussed. For instance, the propagation of unsatisfied water demands, an essential element in the operating rules of dams, occurs at faster time scales than geophysical process but it should not destabilize the simulated waterflows.

B211POSTER-0048: Irrigation enhances local warming with greater nocturnal warming effects than daytime cooling effects

Xing Chen1, Su-Jong Jeong1

1Southern University of Science and Technology (SUSTech), Shenzhen, China

To meet the growing demand for food, land is being managed to be more productive using agricultural intensification practices, such as irrigation. Understanding the specific environmental impacts of irrigation is a critical part of using irrigation as a sustainable way to provide food security. However, our knowledge of irrigation effects on climate is still limited to daytime effects. This is a critical issue in irrigation as a method for mitigating climatic warming related to greenhouse gases (GHGs). This study shows that irrigation led to an increasing temperature (0.005 ℃/year) by enhancing nighttime warming (0.016 ℃/year) more than daytime cooling (-0.007 ℃/year) during the dry season from 1961 to 2014 over the North China Plain (NCP), which is one of largest irrigation areas in the world. By implementing irrigation processes in regional climate model simulations, the consistent warming effect of irrigation on nighttime temperatures over the NCP was shown to match observations. The intensive nocturnal warming is attributed to the energy storage in the wetter soil during the daytime, which contributed to the nighttime surface warming. Our results suggest that irrigation could amplify the warming related to GHGs, and this effect should be explicitly integrated into future climate change projections.

B212 POSTER-0253: Improvement of the modelled infiltration and surface runoff for flash flood events with the JULES land surface model

Chloe Largeron1, Anne Verhoef2, Hannah Cloke3

1University of Reading, Department of Geography and Environmental Science, Reading, United Kingdom 2 University of Reading, Department of Geography and Environmental Science, Reading, United Kingdom 3University of Reading, Department of Geography and Environmental Science and Department of Meteorology, Reading, United Kingdom

Intense rainfall can lead to flash flooding and may cause disruption, damage and loss of life. Since flooding from intense rainfall (FFIR) occurs usually during a short duration and in a limited area, these events are generally poorly predicted by Numerical Weather Prediction (NWP) models, because of the high spatio-temporal resolution required and because of the way the convective rainfall is described in the model. Moreover, the hydrological process descriptions of land surface models are not necessarily suitable to deal with cases of intense rainfall.
In the framework of the TENDERLY project, we improved the representation of the infiltration of the soil in JULES by introducing a variable maximum infiltration. Different scheme of maximum infiltration have been tested to allow us to better predict the amount of surface runoff which is related to the river discharge.
Here, we present the behaviour of infiltration and surface runoff in test cases of intense rainfall with a variable maximum infiltration. This also relates to an adaptation of the modelled hydraulic conductivity profile with depth and with the soil moisture content. The new infiltration scheme is then applied using PDM and the river flow model (RFM) routing scheme. This configuration is being applied to the small Ure catchment and will be tested during a case of flash flooding in the Wansbeck catchment.

B213POSTER-0369: Rainfall trends and mixed fortunes for Botswana's barns

Reason Machete 1, Khumo Semang 1, Nnyaladzi Batisani 1

1Botswana Institute For Technology Research and Innovation , Gaborone , Botswana

Climate change is expected to impact livelihoods and it is important to quantify such impacts. Dryland crop farming is among the livelihoods that are susceptible to climate change impacts. Due to the anticipated rainfall and temperature variability, crop yields of cereal such as maize and sorghum may vary with time. Variability of yield for such dry land arable farming makes it hard for nations to plan ahead for years of want. This problem could be alleviated if index-based crop insurance could be provided. An understanding of past yield trends and rainfall trends could also be important if employed as proxies of future projections. Here we focus on two regions in Botswana and assess their crops and rainfall trends over the period from 1979-2003.We explore the possibility of using annual rainfall as a proxy for yield of the dry-land crop. The results shows narratives and mixed fortunes. Associated uncertainties are quantified to help inform decision making, but what are the implications to food security?