Increasing cloud water resource in a warming world

To understand the regional characteristics of CWR, the Köppen–Geiger climate classification is applied. Global land is divided into five regions, namely, tropical, arid, temperate, cold, and polar [18]. To some extent, the influence of different climatic conditions can be examined by classifying the continents into different climatic types and comparing the variation of CWR in each climatic category. The 2 m temperature and total precipitation of European Centre for Medium-Range Weather Forecasts Reanalysis v5 (ERA5) are used for Köppen classification during 2000–2017 (figure 1(c)).

Our method to quantify CWR is called the 3D time-varying cloud water diagnostic quantification method of CWR based on satellite and reanalysis data [19, 20]. The diagnostic results are mainly based on hydrometeor field products, which are established using CloudSat/CALIPSO satellite data by identifying the relative humidity threshold of cloud area and the profile of cloud water content with temperature. Hydrometeor field products have a total of 21 layers ranging from 1000 hPa to 100 hPa.

For an area defined by a 1° × 1° grid box over one year, CWR is defined as:

$$\begin{equation}{\text{CWR }} = {\text{Mh}}1{ } + {\text{Qhi }} + {\text{Cvh }} - {\text{Ps}}\end{equation} \tag{ 1 }$$

where, Mh1, a state term, is the hydrometeor content at the beginning time in computation period terms, which refers to 0000UTC on 1 January of each year here. Since the state term varies greatly with time, the mean mass of hydrometeor (MMh), monthly mean hydrometeor content, over the study period are often studied for long-term research in our study. It is obtained by vertical integration of the hydrometeor field which is the same as the cloud water path often studied in previous work [21, 22]; Qhi, an advection term, is the total input hydrometeor inflow through the four horizontal boundaries of the 1° grid box. The hydrometeor inflow is calculated using the winds field from National Centers for Environmental Prediction/National Center for Atmospheric Research reanalysis data pointing into the inner grid on the boundary times the hydrometer field. The values at each pressure layer are calculated and then the vertical integration is performed; Cvh, a source term, is the amount of atmospheric hydrometeor converted from water vapor through condensation/deposition processes, a diagnostic term estimated by the equation of atmospheric hydrometeor budget; and Ps is a sink term, the total precipitation deriving from the 1° × 1° daily rainfall product of Global Precipitation Climatology Project. Both advection and source/sink terms are the total annual amount for the unit grid cell. The sum of CWR and Ps is called the gross mass of hydrometeors (GMh). In a certain area for a certain period of time, the calculation method of the gross mass of water vapor (GMv) is similar to that of CWR, and its component terms are atmospheric water vapor at the initial time, water vapor horizontal flow, evaporation from the cloud, and surface evaporation. The evaluation and analysis of CWR and other terms are both carried out based on the average over each 1° grid cell accumulated over each year.

We study the characteristics of potential CWR and its component terms at the scale of a 1° grid over each Köppen zone (figure 1(c)) and the global land area during 2000–2017. The distribution of the vertical integral of CWR (figure 2(a)) is concentrated in the middle-high latitudes of the Northern Hemisphere, especially in Eastern North America and Western Europe. The equator, along with some South Asia regions, also have higher concentrations than surrounding regions, indicating that there are abundant cloud hydrometeors with the potential to turn to precipitation. In terms of the Köppen classification, the cold zone has the most abundant CWR with a mean grid value of 2.78 × 10³ mm (table 1). The amounts of CWR in tropical and temperate zones are 1.50 × 10³ mm and 1.76 × 10³ mm respectively, while the arid zone has the lowest amount of 0.86 × 10³ mm. It is perhaps contrary to the expectation that the tropical zone which is known to be rich in moisture like the mean mass of hydrometeor (MMh, figure 2(b)) or surface precipitation (Ps, figure 2(e)), is not the region with the most abundant CWR. Both CWR and Ps are less abundant in the arid zone. The hydrometeor precipitation efficiency (PEh), the ratio of Ps to the gross mass of hydrometeors (GMh), is high in tropical and temperate zones and low in others (figure 2(f)). The global PEh is about 30.5% over land, of which as high as 53.4% corresponds to the tropical zone. Hence high Ps in the tropics reduces CWR since most hydrometeors contribute to rainfall. Our study shows that a region with the richest moisture substances does not have the greatest unrealized precipitation potential, whereas Western Europe, Northwest North America, Northeast China, and other similar regions have the opportunity to retrieve CWR in the atmosphere as a resource [23, 24].

Zoom In Zoom Out Reset image size

In response to global warming, CWR is rapidly increasing. Not only those regions that are already rich in CWR show strong increases such as Western Europe and North America, but also some water-stressed regions with scarce CWR are also experiencing rapid growth including the Sahara Desert, Western Australia Desert, the Indian Peninsula, the Tibetan Plateau and the Mongolian Plateau (figure 3(a)). The trend distribution of CWR is roughly similar to that of MMh (figure 3(b)). CWR shows an overall increasing trend of 2.33 × 10² mm decade⁻¹, among which the cold zone is increasing at 3.18 × 10² mm decade⁻¹. The tropical zone, due to a decrease near the Amazon Plain and the Congo Basin, shows an overall statistically insignificant growth trend of 1.51 × 10² mm decade⁻¹. Compared with CWR, Ps increases only slowly or even decreases in some regions (figure 3(e)), resulting in a decrease of PEh over land (figure 3(f)). From the perspective of potential atmospheric water resources, increasing CWR on a global scale may increase the potential precipitation resource to be utilized in the future.

Zoom In Zoom Out Reset image size

Since the magnitude of MMh is small (so is the hydrometeor content at the beginning moment (Mh1) (see section 2)), three orders less than CWR, variations in CWR are mainly determined by hydrometeor inflow (Qhi), condensation in the cloud (Cvh) and surface precipitation (Ps). The distributions of CWR vs Qhi (figure 2(c)), as well as Cvh (figure 2(d)) vs Ps are similar with spatial similarity coefficients exceeding 0.99 (p < 0.01) for both pairs. In addition, similar spatial trend patterns of CWR vs Qhi (figure 3(c)), as well as Cvh (figure 3(d)) vs Ps (figure 3(e)), can be demonstrated by highly correlated temporal anomalies, which further demonstrate their pairwise relationships. In other words, for one year within a 1° grid box, most of the hydrometeors that fall to the ground forming precipitation (Ps) come from the condensation process in the cloud (Cvh), while the remaining hydrometeors (CWR) with potential to form precipitation that suspends in the atmosphere are mainly determined by the horizontal hydrometeor flow (Qhi). As the advection term, changes in Qhi are governed by both dynamic and thermodynamic components [25–27], that is to say, fluctuations in atmospheric circulation and local hydrometeor growth.

Global warming is indisputable, with land warming of about 1.59 °C in the last decade compared to pre-industrial times [28]. The main effect of temperature changes on hydrometeor growth can be explained in two ways. One is that a warmer environment is favorable for water vapor storage, following the Clausius–Clapeyron scale [7, 29]. An abundance of water vapor is the necessary condition for the formation of hydrometeor and precipitation, which is consistent with the view that warming speeds up the hydrological cycle [30, 31]. Another is that increased temperature will decrease the stability of the boundary layer conducive to the formation of cloud hydrometeors [32, 33]. In high latitudes, most warming is expected to be focused at the surface (due to snowmelt) and so this may destabilize the boundary layer. This is consistent with the largest increases in CWR being in higher latitudes. It is noted that CWR is the most abundant and fastest-growing in the cold zone, with the strongest trends mainly distributed in the middle-high latitudes near the Arctic and the North Atlantic where temperature and precipitable water both increase significantly (figures 4(a) and (b)). Warming is most pronounced in the Arctic, from two to four times the global average in models [28, 34], known as the Arctic amplification [35, 36]. Arctic warming has contributed to the dramatic melting of Arctic ice. Rapid loss of sea ice, leading to prolonged periods of ice-free seas, particularly in the Barents and Kara seas [37], has accelerated evaporation and increased atmospheric water vapor content accordingly. This has led to an increase in low clouds at the margins of the Arctic ocean [38], which can then be transported southwards to high latitude continents by the polar easterly winds [4, 39].

Zoom In Zoom Out Reset image size

The global atmospheric circulation controls the unstable regions of the atmosphere, the tracks of cyclones and the direction of large-scale flows, determining the horizontal transport path of water vapor, and cloud hydrometeors [40, 41]. The difference in temperature between the Arctic and mid-latitudes is also a fundamental driver for large-scale weather systems [42]. The reduced meridional temperature gradient could broaden the atmospheric ridge and increase the south-north amplitude of the mean flow, which in turn could admit more intense cyclones with moisture resources through a more varied storm track to interior areas of mid-high latitude continents [43, 44]. Arctic amplification tends to encourage the negative phase of Arctic Oscillation/North Atlantic Oscillation and polar westerly jet stream with larger meanders [45, 46]. Such conditions are favorable for the maintenance of persistent pressure anomalies like the Ural blocking high pressure [43]. In this case, poleward flows carry Atlantic hydrometeors northward through Western Europe while equatorward flows carry hydrometeors formed over warming Barents and Kara Seas southward to the heart of the Eurasia continent. Both Western Europe and Eastern North America can receive moisture from the Atlantic if synoptic conditions are favorable, being relatively flat and free from mountain barriers. Hence onshore flow can carry warm and wet water vapor or clouds from the Atlantic deep into the continental interiors.