by Tianjia Liu
cover image by Elayne Fivenson

A new normal in an intensifying global water cycle

While humans have long adapted to regimes of water scarcity or excess, we are underprepared for extreme events — the “mega” droughts, storms, and floods that used to occur once in a hundred or thousand years. But in this decade alone, we witnessed hot, cracked earth in California and flooded fields in the U.S. Midwest (Figure 1), drowned towns in Mozambique and parched wells in Maharashtra, India, and catastrophic storms such as Hurricanes Sandy and Maria. Precipitation extremes — and in particular, rapid see-saw changes from dry to wet spells — severely damage crops, destroy infrastructure, and destabilize livelihoods. But this is our planet’s new normal, and we must adapt.

Figure 1: Flooding inundates agricultural fields along the Mississippi River in Spring-Summer 2019. This pair of satellite images from Landsat 8 show the extent to which the Mississippi River swelled after a flood event. The left panel is from June 5, 2018, and the right panel is from May 7, 2019. (NASA Earth Observatory)

Observing water from space: advancements and challenges

As extreme weather events become more common, we must improve our capabilities to predict their occurrence and closely monitor how they evolve in near real time. How well we do this affects policy efforts to minimize economic cost and protect communities.

Scientists can monitor and predict individual weather events, such as hurricane tracks, at a regional scale and on short timescales (up to ~3-5 days ahead of an event). However, uncertainty quickly increases as projections of how these events evolve extend further and further into the future. On longer timescales, scientists cannot predict exactly when or where the next Category 5 hurricane will strike. Nevertheless, we have a good idea of how the frequency and intensity of extreme events will change given projections of climate and human activity.

The key to improving monitoring and prediction is high-quality observations, which include those of rainfall, air temperature, sea surface temperature and salinity, surface water area, soil moisture, and wind speed and direction. Aside from the traditional sparse network of weather gauges and ocean buoys to track these variables globally, scientists increasingly rely on state-of-the-art satellite observations to fill in observational gaps of these quantities.

For instance, images from Landsat, a series of U.S. Geological Survey satellites that have observed the land surface since the 1970s, can capture the extent of individual flood events at high spatial resolution. Imaging instruments on Landsat platforms function like super-cameras, recording data not only in visible but also in infrared and thermal wavelengths. Access to multiple wavelengths allows scientists to more easily distinguish water from clouds and land features, such as soil, burn scars from wildfires, and buildings. These data-rich images allow for the detection of abnormal features that mark extreme weather events, such as water where there should be dry land, indicating a flood.

In contrast, monitoring droughts is more complex and indirect, as the moisture content in underground soil layers must be considered. One method relies on detecting anomalies in Earth’s gravitational field (“gravity anomalies”) to track total freshwater storage on land. Physically, gravity anomalies reflect abnormal changes in mass — in this case, water. Currently, NASA’s twin Gravity Recovery and Climate Experiment (GRACE) satellites work together to measure gravity anomalies: the relative distance between the two satellites change as they accelerate and decelerate over positive (e.g. water excess) and negative (e.g. water deficit) gravity anomalies. The magnitude of GRACE gravity anomalies can thus reflect the intensity of drought and groundwater depletion (Figure 2).

While satellites have been around for decades, the technology to obtain high quality data from space is only getting better. Advancements in spacecraft technology have led to an exponential increase in the volume of satellite data we collect. This has allowed scientists to model extreme events at finer and finer spatial and temporal scales. Now that there’s a wealth of satellite and weather station data, the next challenge is to develop methods and technology to quickly process and transform the data into actionable policy.

Figure 2: Changes in water storage over land from 2002-2016, as observed by the GRACE satellites. Recent groundwater depletion in California and north India is highlighted. Data are from Rodell et al. (2018).

Regional consequences of drought and floods on agricultural systems

Aside from monitoring and prediction, scientists are trying to better understand the direct and indirect effects of water extremes. One major consequence is the direct stress on agriculture, leading to shifts in food production.

During the multi-year drought in California from 2011-2017, the acres of unused land increased due to water shortage for irrigation. Some farmers switched to less water-dependent crops, such as transforming citrus orchards into vineyards. In contrast, the U.S. Midwest experienced massive flooding along the Missouri, Platte, and Mississippi Rivers in Spring-Summer 2019. As agricultural fields filled with water, crops drowned in waterlogged soils. Additionally, water-driven changes in agricultural systems can have wide-ranging impacts beyond our food, such as negative effects on air quality.

As seen in GRACE data, north India has experienced severe groundwater depletion in recent years, a problem compounded by monsoon rainfall deficits. This problem emerges because rising agricultural production demands heavy groundwater use for irrigation. To address this problem, the agricultural states of Punjab and Haryana enacted a state-wide policy in 2009 that delays the sowing of the rice, a water-intensive crop, to be closer to the onset of the rainy season (from June-September).

However, this may inadvertently worsen north India’s toxic air pollution during the post-monsoon period from October to November. After rice harvests, some farmers openly burn crop residues to clear fields for planting the winter crop, primarily wheat. By shifting the rice growing season, peak post-monsoon burning has also shifted later — by about two weeks — from 2003-2016. As a result, days in the burning season may increasingly coincide with slower winds and colder temperatures – meteorological conditions that favor smog and haze formation. In New Delhi, low visibility resulting from thick haze has led to car crashes, school closings, and airport shutdowns.

A grim picture: rapid changes on the horizon in high latitudes

Drought can also exacerbate fires, especially on carbon-rich peatlands where there’s plenty of fuel for fire. Peat is partially decayed plant matter that slowly accumulates over hundreds to thousands of years in bogs. Across far north regions such as Alaska and Siberia, the rapid thawing of permafrost, or frozen subsurface soil layers, is exposing vast stores of peat.

Under drought conditions, the top layers of peat dry out. Dry peat can fuel smoldering fires that burrow deep into the ground and are hard to extinguish. Peat fires can burn for a long time and release huge amounts of smoke into the atmosphere.

This summer, the Arctic Circle has seen unprecedented levels of fire activity (Figure 3). Russian cities are also suffering from smoke plumes originating from Siberian forest fires. There is uncertainty on how interactions between drought-induced peat fires, increased greenhouse gas emissions, and rapid warming in high latitudes will play out, but the triple threat of warming, drought, and wildfires could be disastrous for ecosystems and human populations living within the Arctic Circle and beyond.

Figure 3: Evidence of fire in the Arctic Circle. Black carbon, or soot, is emitted from wildfires in the Arctic Circle in July 2019 and lofted high into the atmosphere. (NASA Earth Observatory)

Treading water in efforts to adapt to extreme events

Currently, we are vastly underprepared for the direct and cascading effects of extreme water-related events in the upcoming decades. Agricultural systems, natural ecosystems, infrastructure, and livelihoods are all at risk.

We tend to allocate more funding for disaster relief than prevention. This short-term focus on post-disaster recovery is not cost-effective and cannot continue in a world of extremes. For example, the U.S. Federal Emergency Management Agency (FEMA) committed $8.3 billion in disaster mitigation from 2007-2016. However, 84% of this spending, called the Hazard Mitigation Grant Program, is only accessible to individual states after a federal major disaster declaration and comprises only up to 15-20% of total disaster aid.

But there is good news. Research shows that, on average, for every $1 spent on disaster mitigation, taxpayers save $6 on future disaster relief (Figure 4). We can also synthesize long-term adaptation strategies to maximize co-benefits. Wetland restoration, for example, can protect coastal communities by absorbing large amounts of water from storm surges like a sponge. By acting as a buffer, wetlands can lessen the impact of extreme hurricanes, such as Hurricane Sandy and Katrina. Wetlands are also biodiverse ecosystems and carbon sinks that remove carbon dioxide from the atmosphere and store it as soil carbon.

Figure 4: Disaster mitigation vs. prevention. Research shows that for every $1 invested on disaster mitigation, taxpayers save on average $6 on future disaster relief.

Looking ahead, advancements in satellite technology and data collection will continue to improve the scale at which we can monitor and predict floods, storms, and droughts. Recent extreme events remind us that as our planet warms, we are treading water in efforts to mitigate devastation to our infrastructure, livelihoods, and agricultural systems. A paradigm shift from disaster relief to prevention will be crucial for buffering the worst effects of climate change.


Tianjia (Tina) Liu is a third-year Ph.D. student in the Atmospheric Chemistry Modeling Group in the Department of Earth and Planetary Sciences at Harvard University. Her research focuses on fires and air quality in India and Indonesia. Follow her on Twitter @TheRealPyroTina.

Elayne Fivenson is a third-year PhD student in the Biological and Biomedical Sciences program at Harvard Medical School, where she is studying the genetics and biochemistry of the bacterial cell envelope.

For More Information:

This article is part of our special edition on water. To read more, check out our special edition homepage

Leave a Reply

Your email address will not be published. Required fields are marked *