Dear Author: Thank you for publishing with AGU. Your ...

Dear Author:

Thank you for publishing with AGU. Your proofs are attached. Following the annotation guidelines and returning the proofs quickly will allow prompt online posting of the final version of your paper. We have added several new features to the journal Web sites and additional services that we want to bring to your attention. The online version of your paper is enhanced in several ways, and will now include “Altmetrics,” which continually tracks and links to mentions in news outlets, twitter, blogs, and other social media. A link to scholarly citations is also provided. Papers are also available in a regular PDF and as a Readcube version, which includes the Altmetrics and dynamic links to references and figures. Please follow our social media sites! Several of our journals have social media sites; these and other AGU sites are indexed here. If you have not already, we encourage you to create an ORCID (Open Researcher and Contributor ID) so that your work can be uniquely tracked. Your ORCID can be created and linked through your GEMS account. Finally, your feedback is important to us. If you have questions or comments regarding your AGU Publications experience, including information on production and proofs, please contact us at [email protected]. We will also be contacting you soon after your article is published with an author survey. Please take a few minutes to respond to this online survey; your input is important in improving the overall editorial and production process. Thank you again for supporting AGU. Sincerely, Brooks Hanson Director, Publications AGU [email protected]

PUBLICATION FEE CONSENT FORM Dear Author, As documented on the Steps to Publications webpage, content accepted for publication in an AGU journal may be subject to publication fees. The current fee schedule for AGU journals is available from the Author Resources webpage. This consent form has been generated for you in accordance with this policy on the basis of the final word, table, and figure counts of your article. The word count is for the abstract, body text, and captions only. Please review the calculated publication fee below, complete any relevant fields necessary for us to process your final invoice, and return a signed copy of this form to the production editor with your proof corrections. A final invoice will be mailed to the named billing contact within a few weeks of your articles appearance online in edited format. Please note: OnlineOpen, the open access option for AGU journals, must be purchased through the OnlineOpen order form (https://authorservices.wiley.com/bauthor/onlineopen_order.asp). If you elected OnlineOpen at any earlier phase of submission and have not already submitted your order online, please do so now. Authors who select OnlineOpen will not be charged any base publication fee, but excess length fees will still apply. ARTICLE DETAILS Journal: Geophysical Research Letters Article: Devineni, N., U. Lall, E. Etienne, D. Shi, and C. Xi (2015), America’s water risk: Current demand and climate variability, Geophys. Res. Lett., 42, doi:10.1002/2015GL063487. Words: 3,518 Figures: 4 Tables: 0 OnlineOpen: No

Publishing Units: 11 Excess Publishing Units: 0 Publication Fee: USD $500

BILLING DETAILS Name: ____________________________________ Address: __________________________________ ___________________________________ ___________________________________

Phone: ____________________________________ E-mail: ____________________________________

Signature: _________________________________

Date: _____________________________________

Journal Code G R L

Article ID Dispatch: 24.03.15 5 2 7 5 3 No. of Pages: 9

CE: DOBT ME:

PUBLICATIONS Geophysical Research Letters RESEARCH LETTER 10.1002/2015GL063487 Key Points: • New water risk indicators are presented • Spatial competition between counties and water storage required is exposed • Future scenarios for water trading and value of water can be developed

Correspondence to: N. Devineni, [email protected]

Citation: Devineni, N., U. Lall, E. Etienne, D. Shi, and C. Xi (2015), America’s water risk: Current demand and climate variability, Geophys. Res. Lett., 42, doi:10.1002/ 2015GL063487.

America’s water risk: Current demand and climate variability Naresh Devineni1, Upmanu Lall2, Elius Etienne1, Daniel Shi3, and Chen Xi4 1

Department of Civil Engineering, NOAA-Cooperative Remote Sensing Science and Technology Center, City University of New York, New York, New York, USA, 2Department of Earth and Environmental Engineering, Columbia Water Center, Columbia University, New York, New York, USA, 3Department of Economics and Mathematics, Columbia College, Columbia University, New York, New York, USA, 4State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Hohai University, Nanjing, China

Abstract A new indicator of drought-induced water stress is introduced and applied at the county level in USA. Unlike most existing drought metrics, we directly consider current daily water demands and renewable daily water supply to estimate the potential stress. Water stress indices developed include the Normalized Deficit Cumulated to represent multiyear droughts by computing the maximum cumulative deficit between demand and supply over the study period (1949–2009) and the Normalized Deficit Index representing drought associated with maximum cumulative deficit each year. These water stress indices map directly to storage requirements needed to buffer multiyear and within-year climate variability and can reveal the dependence on exogenous water transferred by rivers/canals to the area. Future climate change and variability can be also incorporated into this framework to inform climate-driven drought for additional storage development and potential applications of water trading across counties.

Received 13 FEB 2015 Accepted 11 MAR 2015 Accepted article online 16 MAR 2015

1. Introduction Competition for water across different water entities has been identified lately as many regions have experienced a series of droughts in the United States (U.S.). This has increased the awareness of climate-related water risks for agriculture producers, municipalities, energy companies, and business enterprises. Business and investment decisions in infrastructure seek information as to the potential risk of water shortages that may be faced by the investment and the associated cost of such a shortage. Especially in the Colorado River Basin, where the spatial competition for water between upstream and downstream users has increased due to population growth and economic development, questions as to the viability of the endogenous water resources in each county, the value of storage for water, and the value of trading water across the region have emerged. As competition for water increases across different water sectors, the temporal variability of available water supply leads to increasing pressure to develop surface storage, or to use groundwater resources. Given increasing water scarcity and the associated deterioration of the quantity and quality of water sources in many parts of the world, many “tools” have been developed to represent potential water risks. These assessments or tools use global hydrologic models to estimate the distribution of runoff (renewable freshwater) and various water use metrics to project the current and future distributions of water deficits globally [e.g., Wallace, 2000; Oki and Kanae, 2006; Reig, 2013]. Efforts have also been made to analyze water scarcity at more refined spatiotemporal scales both globally and nationally [Alcamo et al., 1997; Meigh et al., 1998; National Integrated Drought Information System, 2007; Vörösmarty et al., 2000], while implication of drought management is also investigated at regional and local scales [Ryu et al., 2014]. They also provide a useful first look, such as a functional relationship between water supply (surface water and groundwater resources) and demand at basin scales. Since they include river flow, and account only for existing abstractions of river flow by current usage, they implicitly consider that renewable water in a county can be transferred to potential downstream users without considering the potential competition for that resource between upstream and downstream users. Typically, these tools use estimates of the average water supply and demand at the subbasin level. Consequently, they understate the potential water risk due to climate variations. In other words, the precipitation intermittence and amount distribution will generally lead to a higher estimate of stress than if just the average precipitation and average demand are used to compute stress. In most places, even if the

DEVINENI ET AL.

©2015. American Geophysical Union. All Rights Reserved.

1

1 2 3 4 5 6 7 8 9 10 11 Q2 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 Q3 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67

1 2 3 4 5 6 resource is not over appropriated on average, persistent shortage induced by climate conditions can lead to 7 stress. A detailed review of the current water scarcity indices can be found in Brown [2011]. 8 Unlike past work that considers estimates of groundwater recharge and river flow as measures of supply [Oki 9 and Kanae, 2006; Alcamo et al., 1997], we use local precipitation in each county to estimate the renewable 10 11 water supply endogenous to the area and consider natural and human uses of this water. The reliance on imported river water or mined groundwater is exposed in the process. This is important to establish in the 12 13 face of spatial competition for existing water resources. Furthermore, to properly diagnose water risk, one needs to examine both existing demand and variations in renewable water supply at an appropriate spatial 14 resolution and unit. A metric that can inform the potential severity of a shortage is the accumulated deficit 15 16 between demand and supply at a location. Here we provide estimates of this metric considering within 17 year and multiyear climate variations and map the resulting risks for the continental USA at a county level 18 using current estimates of water use and a 61 year climate record (1949–2009). 19 20 2. Data Description and USA Context 21 22 Gridded daily rainfall and temperature data from 1949 to 2009 (61 years) available at 1/8° by 1/8° spatial 23 resolution [Maurer et al., 2002] were interpolated to each of the 3111 counties in the continental USA. For 24 computing the agricultural water demands, the most recent data (available for year 2010 across the USA) on 25 47 harvested crops and the total cropland for each county were extracted from the National Agricultural 26 Statistics Services (NASS, http://www.nass.usda.gov/). The daily crop water requirements are estimated based 27 on Food and Agriculture Organization recommended crop coefficients and reference crop evapotranspiration Q4 28 based on Penman-Monteith equation [Allen et al., 1998]. Estimates for the county level industrial, livestock, 29 30 mining, aquaculture, and thermoelectric and domestic water withdrawals were obtained from United States Geological Survey (USGS) water use database for 2010 (http://water.usgs.gov/watuse/). To review the patterns 31 32 of annual rainfall across the U.S., the average annual precipitation and its coefficient of variation (standard deviation divided by mean) are shown in Figures 1a and 1b, respectively. The coastal regions and the Northeast F133 appear to be well endowed with precipitation, while the Southwest and parts of the Midwest are marked by 34 35 high variability in precipitation across years. The interior West is dry. The spatial pattern associated with cropped 36 area is shown in Figure 1c, and the total estimated annual water demand by county is presented in Figure 1d. 37 Not surprisingly, agricultural water use is dominant. 38 39 3. Methodology 40 We developed two risk metrics to capture the effect of within-year dry periods (Normalized Deficit Index, NDI) 41 42 and of drought over multiple years (Normalized Deficit Cumulated, NDC). The Water Risk Indices presented 43 here (see also Devineni et al. [2013] and Chen et al. [2013]) are based on the sequent peak algorithm originally 44 developed for reservoirs. It quantifies the water storage capacity needed to meet the demand for a given 45 sequence of supply [Lall and Miller, 1988; Loucks et al., 1981; Thomas and Burden, 1963]. 46 The steps for the computation of these two indices are presented below. For the jth geographical unit, define 47 48 the following quantities: 49 50 deficitj;t ¼ max deficitj;t1 þ Dj;t Sj;t ; 0 where deficitj ;t¼0 ¼ 0 (1) 51 52 SICj ¼ maxt deficitj;t ; t ¼ 1 : n365 (2) 53 Q7 54 55 SIIjy ¼ maxt deficitj;t ðy Þ ; t ¼ 1 : 365; y ¼ 1 : n ; where deficitj ; t¼0ðyÞ ¼ 0; y ¼ 1…n (3) 56 57 58 wherein deficitj,t refers to the accumulated daily deficit, Dj,t to total daily water demand, Sj,t to the total water 59 daily supply volume, for geographical location j, and day t; y refers to a calendar or cropping year; and n is the 60 total number of years in the analysis. The maximum accumulated deficit estimated over the n year period 61 without breaking it into subperiods is defined as SICj (Stress Index Cumulated). This measures the potential 62 impact of multiyear droughts. For an n year record, intra-annual water stress is evaluated as the maximum 63 64 cumulative deficit each year is defined here as SIIjy (Stress Index Intra-annual). 65 66 ©2015. American Geophysical Union. All Rights Reserved. 2 67

Geophysical Research Letters

DEVINENI ET AL.

10.1002/2015GL063487


10.1002/2015GL063487

Figure 1. (a) Mean annual precipitation in mm/yr, (b) coefficient of variation of annual precipitation in %, (Source: Maurer et al. [2002]) (c) 2010 U.S. net cropped area in county/county area, and (Source: NASS) (d) average annual water demand Mgal/d for the USA (estimated).

The corresponding normalized indices are simply NDCj ¼

SICj ; APj

NDIjy ¼

SIIjy APj

(4)

where APj is the average annual rainfall volume (county area × average depth of precipitation) for county j. The renewable water supply Sj,t is estimated as follows: Sj;t ¼ αj Pj;t AC j þ βj Pj;t Aj ACj

(5)

where Pj,t is the rainfall for any day t, over a county j. Aj is the geographical area of the county, and ACj is the net cropped area for the county. αj and βj are the factors that determine the usable fraction of rainfall for irrigation and for other uses. For this analysis, αj is estimated based on the long-term runoff ratio of the R county. The long-term runoff ratio (Pjj ) (an index computed to understand the partitioning of rainfall into runoff and evaporation) is often related to physiographic basin features and regional climate information R [Berger and Entekhabi, 2001]. We estimate αj as 1 Pjj to reflect the average fraction of rainfall that remains for the consumption of the crops after runoff. For this study, we obtain the long-term runoff ratio based on Sankarasubramanian and Vogel [2003]. We chose βj, which is the fraction of rainfall that can be harvested and utilized from the noncropped area (Aj ACj) to be 0.1. The parameters αj and βj conceptually embody the DEVINENI ET AL.


3

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 Q5 Q6 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 Q8 63 64 65 66 67

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 Figure 2. Magnitude of water stress considering 1949–2009 daily precipitation and 2010 water demand data across the 54 continental USA for (a) within-year NDImax (the worst annual deficit in 61 years) and (b) multiyear cumulative deficit NDC. 55 56 57 processes one could model for bare soil evaporation, soil moisture dynamics, and runoff generation. They can 58 59 be varied parametrically if needed to assess the sensitivity of the final results to the assumed values. 60 The daily water deficit is defined as the difference between the daily water demand and the daily renewable 61 water supply. The deficits are accumulated (equation (1)). The maximum accumulated deficit in a given year 62 divided by the average annual rainfall across the historical period is the NDI for that year. Similarly, the NDC is 63 the maximum accumulated deficit for all 61 years divided by the average annual rainfall. The NDI is computed 64 65 66 ©2015. American Geophysical Union. All Rights Reserved. 4 67


DEVINENI ET AL.

10.1002/2015GL063487

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Figure 3. Spatial categorization of the magnitude and distribution of water deficits and drought risks in USA. Blue: Multiyear 33 Stress (NDC) = Worst Single-Year Stress (NDImax). Orange: NDC > NDImax. Red: Demand exceeds average annual endogenous 34 35 supply in the county. 36 37 38 as one number for each year using historical daily rainfall data for the area and current daily water needs. 39 It measures the maximum cumulated water shortage each year that needs to be provided from additional 40 water resources within the basin or other areas. The deficit at the beginning of each year is set to 0 for the 41 calculation of the NDI, but not for the NDC. Hence, with ~60 years of data, the sixth largest NDI value indicates 42 that there is an approximately 10% chance that water storage or transfers of that amount may be needed to 43 44 meet demands at that location in any given year if multiyear droughts were not considered. The NDC is 45 computed as one number over the historical climate record. It represents the largest cumulative deficit 46 between renewable supply and water use over the entire period. Consequently, it reflects the stress associated with multiyear and within-year shortages at a location. The cumulative deficit measure developed 47 48 here implicitly accounts for the differences between counties in the temporal distribution of rainfall and 49 demand within and across years. By accounting for the timing of planting, different stages of crop growth 50 and the timing and distribution of rainfall and other demands, the model discriminates between two 51 counties that have same total rainfall or demand but differ in that one may have rainfall distributed uniformly 52 over the year through modest rainfall events, while the other may have a few intense rainfall events separated by 53 the long dry periods. The latter gives rise to much higher deficit. A detailed description of the mathematical 54 model along with applications and interpretation can be found in Devineni et al. [2013]. 55 56 57 58 4. Results 59 We first present maps of the NDImax (the worst annual deficit, NDImax) across the 61 years and the NDC in 60 Figure 2. The implications of their difference are then discussed using Figure 3. Finally, trends in the NDI F261 F3 are presented in Figure 4. Since the current demand and the 61 years of climate data are used, these trends F462 allow us to assess how changing climate over this period may have increased or decreased within-year water 63 64 stress at a county level relative to 2010 estimated county demand. 65 66 ©2015. American Geophysical Union. All Rights Reserved. 5 67


DEVINENI ET AL.

10.1002/2015GL063487


10.1002/2015GL063487

Figure 4. Time trends identified through Mann-Kendall trend test on NDI by county (significance level for null hypothesis of no trend = 5%).

An NDI or NDC greater than 1 represents the case where the cumulative deficit is greater than the average endogenous rainfall. The annual rate of consumption in these regions could be higher than the average utilizable rainfall rates. From Figure 2 we see that most of the country has NDImax < 1. As one considers persistence in drought beyond 1 year, we see from the NDC map that the current use patterns portend severe stress over much of the agricultural belt of the Midwest USA as well as the arid regions of California and Arizona. In many parts of the country, a majority of water consumption is dominated by irrigation that relies heavily on nonrenewable groundwater to meet the needs. Chronic or multiyear stress consequently emerges as the event of concern in these areas with many locations requiring greater than 2 to 5 times the average annual rainfall in the location in storage or to be transferred from other locations to make it. A brief case-by-case analysis that identifies the current sources being used by each of the stressed areas highlighted in Figure 2 is provided here. 4.1. Case 1: California Area The Central Valley Project (CVP), the State Water Project (SWP), and the Colorado River form the main water sources for California. The CVP facilitated by up to 20 reservoirs on the Sacramento River, American River, Trinity River, Stanislaus River, and San Joaquin and San Luis Rivers provides up to 11 million acre-feet of water to the Contra Costa Counties and the counties in Santa Clara Valley. The SWP, operated by the California Department of Water Resources, includes 22 reservoirs and 444 mi aqueduct. It is the major source of supply for the cities of Los Angeles, Riverside, San Bernardino, San Diego, and parts of Southern California. In addition, the Colorado River, through its Colorado River Aqueduct, serves as a vital source for urban areas in Southern California. Furthermore, groundwater (which serves up to 38% of annual water supply (http://www.water.ca.gov/ groundwater/index.cfm)) and local reservoirs and water systems also play a major role in water supply for the stressed areas of California. It is interesting to note that much of the water supply in California is based on imported water, as identified by the NDC.

DEVINENI ET AL.


6

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 Q9 56 57 Q10 58 59 60 61 62 63 64 65 66 67


10.1002/2015GL063487

4.2. Case 2: Agricultural Belt of the Midwest and the High Plains Many counties in the Midwest and High Plains have very large NDC values corresponding to average demand exceeding average supply. These counties form the agricultural belt that fall under the most intensively irrigated areas in the country. They are part of the Cambrian-Ordovician aquifer system and the Ogallala aquifer system. In much of these counties, groundwater is widely used for agricultural, municipal, and industrial supplies. Substantial pumping of groundwater for irrigation since the early 1950s led to a large water table decline. The areas identified with high NDC/NDImax also correspond to the maps of the groundwater depletion or over exploitation recently presented by Konikow [2013]. 4.3. Case 3: Metropolitan Areas in the Northeast We also note that metropolitan areas such Washington D.C. and New York City have high NDC values. Using county data and restricting the definition of available supply to precipitation in the measured unit, the methodology effectively exposes the potential problems during deficit years and their reliance on external water transfers. Thus, potential spatial competition between cities and their hinterland can be exposed. For instance, the New York metro area relies on water resources from an upstream watershed (Delaware River Basin) where regional agriculture, water supply, ecology, and development are severely restricted. The cross-sectoral water conflicts during the 1960s drought in the Northeast and the current water compacts [Delaware River Basin Commission, 2007] exposed the endogenous versus exogenous water supply issues in New York. To expose where multiyear drought is an important factor over the USA, we highlight the regions susceptible to persistent drought resulting from natural variations in climate and existing demand in Figure 3, using the ratio of NDC/NDImax. The counties shown in blue have NDC equal NDImax, i.e., multiyear droughts do not have an impact worse than that of the driest year on record. This could either be due to the absence of long droughts or due to a relative level of demand that is low enough to not require storage across years. The counties marked in orange have NDC greater than NDImax , and the ones for which NDC is more than 10 times NDImax are marked in red. In these cases demand reduction may be particularly beneficial unless a high amount of storage or import is available. The red case reflects demands that often exceed total endogenous supply and reflects locations where desalination or groundwater mining or imported water is necessary to meet existing demands. Based on the 61 year (1949–2009) time series of NDI estimates, we assessed the monotonic trends in the incidence of drought events using the Mann-Kendall nonparametric trend test [Mann, 1945; Helsel and Hirsch, 1992]. Figure 4 shows the results from the test for each county. The counties colored red indicate that the NDI has an increasing trend that is statistically significant at the 95% confidence level. Similarly, the counties colored green have a decreasing trend in NDI that is statistically significant at the 95% confidence level. The counties in white have no statistically significant trends. It is important to note that the trend analysis shown here is not the trend on the changing water stress over time. It is the trend one can observe through changing climate forcing induced on current demands. Given the current level of demand, we are interested in showing whether or not the changes in climate translate into an increasing or decreasing water stress relative to current use. Large, nearly contiguous areas with decreasing trends in NDI are observed in the Midwestern and Northeastern United States indicating that the increase in precipitation in these locations directly translates to decreasing stress relative to today’s demand. The spatially contiguous regions with decreasing water stress are in the Dakotas to New England. These regions have experienced an increase in the precipitation in recent times [Environmental Protection Agency, 2013]. An increase in the rainfall directly translates into reduced deficit in the region. Hence, the decreasing trends are consistent with the observed trends in rainfall. Figure 4 also shows that there are a few counties (though sporadic) in southeastern Idaho with increasing droughts and a few counties in southwestern Idaho with decreasing droughts. This is consistent with Sohrabi et al. [2013] analysis. Scattered locations in the Intermountain West and in the Southeast exhibit increasing trends. These are consistent with some recent works on the twentieth century drought trends [Andreadis and Lettenmaier, 2006].

5. Summary and Discussion Many indices of drought, e.g., the PDSI and the SPI, are available in the literature. However, most of these consider only precipitation and/or temperature and averaging over a specific time window and do not consider the attributes of demand. Consequently, they have a limited utility for water resource management.

DEVINENI ET AL.


7

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Q11 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 Q12 52 53 54 55 56 57 58 59 60 61 Q13 62 63 64 65 66 67


10.1002/2015GL063487

On the other hand, measures of water risk or water stress that have been used recently only use time-averaged water supply estimates and hence are not cognizant of drought severity and duration [Reig et al., 2013]. The metrics introduced here address both these issues, working with current demand and a climate scenario. The application presented here only considered the historical climate data and not a future projection but did examine trends in the drought severity across the USA relative to the current demand pattern. An interesting observation is that the dominant trend in many areas of concern is that of a decrease in the water stress metric defined here for within year stress. However, we noted that the multiyear drought metric is actually more important for the USA than the within-year drought stress. Consequently, it is important to develop realistic multiyear climate-induced drought scenarios that have the right spatial structure to analyze the risk of droughts in the current or future climate. These could then provide the basis for the economic valuation of water storage and trading across sectors and locations. For climate informed analyses of water risk, the index emphasizes that the climate information needs to properly represent the time sequence of supply and demand and not just average seasonal or annual values. Where climate models are used for projections of future water supply and its variations, an assessment of their ability to reproduce the temporal structure and seasonality of daily rainfall, and the sequential properties of rainfall across years is important. Since phenomena such as the El Niño-Southern Oscillation play an important role in both seasonal and interannual variations of precipitation [Ryu et al., 2010], the ability to properly model these attributes of the phenomena and their continental teleconnections, as well as how these phenomena are expected to change under anthropogenic forcing, is important for informing future decisions on investments in municipal water infrastructure and in business location across the country. Acknowledgments This research was supported by (a) Veolia Water and Growing Blue, (b) NSF grant 1360446 (Water Sustainability and Climate WSC Category 3), (c) PSC-CUNY award 67832–00 45, and (d) NOAA-CREST-Cooperative agreement NA11SEC4810004. The statements contained within the manuscript/research article are not the opinions of the funding agency or the U.S. government but reflect the author’s opinions. Data used for computing the water stress indices are available from (a) National Agricultural Statistics Survey (NASS: http://www.nass. usda.gov), (b) USGS Water Use (http:// water.usgs.gov/watuse/), and (c) Climate data available at http://www.engr.scu. edu/~emaurer/gridded_obs/index_gridded_obs.html. Data on the water stress indices are available on request from the authors. The Editor thanks two anonymous reviewers for their assistance in evaluating this paper.

DEVINENI ET AL.

References Alcamo, J., P. Döll, F. Kaspar, and S. Siebert (1997), Global change and global scenarios of water use and availability: An application of WaterGAP1.0. Germany: Center for Environmental Systems Research (CESR), University of Kassel. Alcamo, J., P. Döll, T. Hendrichs, F. Kaspar, B. Lehner, T. Rosch, and S. Siebert (2003), Development and testing of the WaterGAP 2 global model of water use and availability, Hydrol. Sci., 48(3), 317–337. Allen, R. G., L. S. Pereira, D. Raes, and M. Smith (1998), Crop evapotranspiration—Guidelines for computing crop water requirements, FAO Irrigation and drainage paper 56. Andreadis, K. M., and D. P. Lettenmaier (2006), Trends in 20th century drought over the continental United States, Geophys. Res. Lett., 33, L10403, doi:10.1029/2006GL025711. Berger, P. K., and D. Entekhabi (2001), Basin hydrologic response relations to distributed physiographic descriptors and climate, J. Hydrol., 247, 169–182. Brown, A. (2011), A review of water scarcity indices and methodologies, University of Arkansas Sustainability Consortium, White Paper # 106 (February 27): 1–21. Chen, X., N. Devineni, U. Lall, H. Zenchun, D. Leihua, J. Qin, W. Jiahu, and W. Shicheng (2013), China’s water sustainability in the 21st century: A climate informed water risk assessment covering multi-sector water demands, Hydrol. Earth Syst. Sci. Discuss., 10, 11,129–11,150. Devineni, N., S. Perveen, and U. Lall (2013), Assessing chronic and climate induced water risk through spatially distributed cumulative deficit measures: A new picture of water sustainability in India, Water Resour. Res., 49, 2135–2145, doi:10.1002/wrcr.20184. Delaware River Basin Commission (2007), Flexible flow management program, September 26, 2007. Environmental Protection Agency (2013[Available at http://www.epa.gov/climatechange/pdfs/precipitation_documentation.pdf.]). Helsel, D. R., and R. M. Hirsch (1992), Statistical Methods in Water Resources, Studies in Environmental Science, vol. 49, 522 pp., Elsevier Science. Konikow, L. F. (2013), Groundwater depletion in the United States (1900–2008), U.S. Geological Survey Scientific Investigations Report 2013–5079, 63 p., [Available at http://pubs.usgs.gov/sir/2013/5079.] Lall, U., and C. W. Miller (1988), An optimization model for screening multipurpose reservoir systems, Water Resour. Res., 24(7), 953–968, doi:10.1029/WR024i007p00953. Loucks, D. P., J. R. Stedinger, and D. A. Haith (1981), Water Resource Systems Analysis, Prentice-Hall, Englewood Cliffs, N. J. Mann, H. B. (1945), Nonparametric tests against trend, Econometrica, 13, 245–259. Maurer, E. P., A. W. Wood, J. C. Adam, D. P. Lettenmaier, and B. Nijssen (2002), A long-term hydrologically-based data set of land surface fluxes and states for the conterminous United States, J. Clim., 15(22), 3237–3251. Meigh, J. R., A. A. McKenzie, B. N. Austin, R. B. Bradford, and N. S. Reynard (1998), Assessment of Global Water Resources e Phase II. Estimates of Present and Future Water Availability in Eastern and Southern Africa, Center for Ecology and Hydrology, Wallingford, U. K. National Integrated Drought Information System (2007), The National Integrated Drought Information System implementation plan, National Integrated Drought Information System. [Available online at http://www.drought.gov.] Oki, T., and S. Kanae (2006), Global hydrological cycles and world water resources, Science, 313, 1068–1072. Reig, P., T. Shiao, and F. Gassert (2013), Aqueduct Water Risk Framework. Working Paper, World Resources Institute, Washington, D. C., [Available at http://www.wri.org/publication/aqueduct-water-risk-framework.] Ryu, J. H., M. D. Svoboda, J. D. Lenters, T. Tadesse, and C. L. Knutson (2010), Potential extents for ENSO-driven hydrologic drought forecasts in the United States, Clim. Change, 101, 575–597. Ryu, J. H., M. Sohrabi, and A. Acharya (2014), Toward mapping gridded drought indices to evaluate local drought in a rapidly changing global environment, Water Resour. Manage., 28, 3859–3869. Sankarasubramanian, A., and R. M. Vogel (2003), Hydroclimatology of the continental U.S., Geophys. Res. Lett., 30(7), 1363, doi:10.1029/ 2002GL015937.


8

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Q14 33 34 35 36 37 38 39 40 41 42 43 44 45 46 Q15 47 48 49 50 51 52 53 54 55 56 Q16 57 58 59 60 61 62 63 64 65 66 67


10.1002/2015GL063487

Sohrabi, M. M., J. H. Ryu, J. Abatzoglou, and J. Tracy (2013), Climate extreme and its linkage to regional drought over Idaho, USA, Nat. Hazards, 65(1), 653–681. Thomas, H. A., Jr., and R. P. Burden (1963), Operations Research in Water Quality Management, pp. 1–17, Harvard Water Resources Group, Cambridge, Mass. Vörösmarty, C. J., P. Green, J. Salisbury, and R. Lammers (2000), Global water resources: Vulnerability from climate change and population growth, Science, 289, 284–288. Wallace, J. S. (2000), Increasing agricultural water use efficiency to meet future food production, Agric. Ecosyst. Environ., 82, 105–119.

DEVINENI ET AL.


9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67

Author Query Form Journal: Geophysical Research Letters Article: grl_52753 Dear Author, During the copyediting of your paper, the following queries arose. Please respond to these by annotating your proofs with the necessary changes/additions. • If you intend to annotate your proof electronically, please refer to the E-annotation guidelines. • If you intend to annotate your proof by means of hard-copy mark-up, please use the standard proofing marks. If manually writing corrections on your proof and returning it by fax, do not write too close to the edge of the paper. Please remember that illegible mark-ups may delay publication. Whether you opt for hard-copy or electronic annotation of your proofs, we recommend that you provide additional clarification of answers to queries by entering your answers on the query sheet, in addition to the text mark-up.

Query No.

Query

Q1

AUTHOR: Please complete the Publication Fee Consent Form included with your article and return to the Production Editor with your proofs.

Q2

AUTHOR: Please confirm that given names (red) and surnames/family names (green) have been identified correctly.

Q3

AUTHOR: “Please provide complete reference details of Reig [2013].

Q4

AUTHOR: "FAO" has been expanded as "Food and Agriculture Organization". Please change if incorrect.

Q5

AUTHOR: Figures 1 to 4 contain small and poor quality text. Please check and resupply if necessary.

Q6

AUTHOR: Please check that all figure captions have been presented correctly.

Q7

AUTHOR: Please advise if convolution is meant by the asterisk in equation 2. Otherwise, please change it to multiplication sign.

Q8

AUTHOR: The citation “Sankarasubramanian and Vogel [2002]” (original) has been changed to “Sankarasubramanian and Vogel, 2003”. Please check if appropriate.

Q9

AUTHOR: Please convert acre-feet value to m3.

Q10

AUTHOR: Please convert values of miles to km.

Q11

AUTHOR: "DRBC" has been expanded as "Delaware River Basin Commission". Please change if incorrect.

Q12

AUTHOR: "EPA" has been expanded as "Environmental Protection Agency". Please change if incorrect.

Q13

AUTHOR: Please define "PDSI and the SPI".

Q14

AUTHOR: Reference “Alcamo et al. (2003)” is not cited in the text. Please indicate where it should be cited; or delete from the reference list.

Q15

AUTHOR: Please provide the city location of publisher for Reference Helsel and Hirsch 1992.

Remark

Query No.

Query

Q16

AUTHOR: "NIDIS" has been expanded as "National Integrated Drought Information System". Please change if incorrect.

Remark

USING e-ANNOTATION TOOLS FOR ELECTRONIC PROOF CORRECTION Required software to e-Annotate PDFs: Adobe Acrobat Professional or Adobe Reader (version 7.0 or above). (Note that this document uses screenshots from Adobe Reader X) The latest version of Acrobat Reader can be downloaded for free at: http://get.adobe.com/uk/reader/ Once you have Acrobat Reader open on your computer, click on the Comment tab at the right of the toolbar:

This will open up a panel down the right side of the document. The majority of tools you will use for annotating your proof will be in the Annotations section, pictured opposite. We’ve picked out some of these tools below:

1. Replace (Ins) Tool – for replacing text.

2. Strikethrough (Del) Tool – for deleting text.

Strikes a line through text and opens up a text box where replacement text can be entered. How to use it

Strikes a red line through text that is to be deleted. How to use it



Highlight a word or sentence.



Highlight a word or sentence.



Click on the Replace (Ins) icon in the Annotations section.



Click on the Strikethrough (Del) icon in the Annotations section.



Type the replacement text into the blue box that appears.

3. Add note to text Tool – for highlighting a section to be changed to bold or italic.

4. Add sticky note Tool – for making notes at specific points in the text.

Highlights text in yellow and opens up a text box where comments can be entered. How to use it

Marks a point in the proof where a comment needs to be highlighted. How to use it



Highlight the relevant section of text.





Click on the Add note to text icon in the Annotations section.

Click on the Add sticky note icon in the Annotations section.



Click at the point in the proof where the comment should be inserted.



Type the comment into the yellow box that appears.



Type instruction on what should be changed regarding the text into the yellow box that appears.

USING e-ANNOTATION TOOLS FOR ELECTRONIC PROOF CORRECTION

5. Attach File Tool – for inserting large amounts of text or replacement figures.

6. Add stamp Tool – for approving a proof if no corrections are required.

Inserts an icon linking to the attached file in the appropriate pace in the text. How to use it

Inserts a selected stamp onto an appropriate place in the proof. How to use it



Click on the Attach File icon in the Annotations section.



Click on the Add stamp icon in the Annotations section.



Click on the proof to where you’d like the attached file to be linked.





Select the file to be attached from your computer or network.

Select the stamp you want to use. (The Approved stamp is usually available directly in the menu that appears).



Click on the proof where you’d like the stamp to appear. (Where a proof is to be approved as it is, this would normally be on the first page).



Select the colour and type of icon that will appear in the proof. Click OK.

7. Drawing Markups Tools – for drawing shapes, lines and freeform annotations on proofs and commenting on these marks. Allows shapes, lines and freeform annotations to be drawn on proofs and for comment to be made on these marks..

How to use it 

Click on one of the shapes in the Drawing Markups section.



Click on the proof at the relevant point and draw the selected shape with the cursor.



To add a comment to the drawn shape, move the cursor over the shape until an arrowhead appears.



Double click on the shape and type any text in the red box that appears.

For further information on how to annotate proofs, click on the Help menu to reveal a list of further options:

Additional reprint and journal issue purchases

Should you wish to purchase additional copies of your article, please click on the link and follow the instructions provided:

https://caesar.sheridan.com/reprints/redir.php?pub=10089&acro=GRL

Corresponding authors are invited to inform their co‐authors of the reprint options available. Please note that regardless of the form in which they are acquired, reprints should not be resold, nor further disseminated in electronic form, nor deployed in part or in whole in any marketing, promotional or educational contexts without authorization from Wiley. Permissions requests should be directed to mailto: [email protected] For information about ‘Pay‐Per‐View and Article Select’ click on the following link: http://wileyonlinelibrary.com/ppv