Introducing WSIM Version 2

22 May 2019

VERSION 2 HIGHLIGHTS
ISciences is transitioning production of our monthly Global Water Monitor & Forecast Watch List reports to use outputs from a new version of our Water Security Indicator Model (WSIM). Version 2 (v2) completely re-engineered WSIM to:

Support a wide variety of use cases,
Make use of more recent and publicly available input datasets,
Be more computationally and operationally efficient, and
Revise the methodology for calculating composite water anomalies.

WSIMv2 has been published as an open-source toolkit that is available on Docker Hub and GitLab, with full documentation and examples published at wsim.isciences.com. It is primarily written in R with selected components written in C++, Python, and bash.

WSIMv2 maintains the central premise of WSIMv1 – that populations structure their activities based upon expected climatic provisions of fresh water, and are able to maintain these activities in response to a certain degree of variation in the amount, frequency, and timing of fresh water provisions. When variation exceeds that experienced in the historical record, however, populations may be forced to react in unexpected ways. While the relationship between climatic stresses and populations is complex, these reactions could induce transnational water disputes, agricultural shortfalls, electricity shortages, population displacements, infectious disease outbreaks, and perhaps political instability. The premise that societies react to unusual changes in the hydrology of their particular location motivates WSIM to characterize water stresses in terms of anomalies – deviations of actual from expected conditions – expressed as return periods.

REPORTING COMPOSITE SURPLUSES AND DEFICITS
A core feature of WSIM is to model aggregate fresh water stresses as composite surplus indices and composite deficit indices, and produce maps depicting their severity as a reflection of their rarity, which is measured by their return periods.

In both WSIMv1 and WSIMv2, the Composite Deficit Index is calculated as the minimum of the return periods computed for each of three indicators: soil moisture, EmPET (actual minus potential evapotranspiration), and total blue water. The Composite Surplus Index is calculated as the maximum of two indicators: runoff and total blue water. WSIMv2 introduces a new method for calculating the return periods for the composite indices.

WSIMv1 reports the return period of a composite index as the value of the index itself. As such, the return period for surpluses represents the maximum of the return periods of the index’s component surplus indicators, and the return period for deficits is the minimum of the return periods of the component deficit indicators. This method may overestimate the rarity of a surplus or deficit by taking the most severe of multiple indicators. WSIMv2 addresses this by additionally producing an adjusted composite return period. The adjusted composite return period calculates the surplus and deficit indices as in WSIMv1, but then calculates the return period with respect to the historical distribution of the composite index itself, rather than merely reflecting the “worst” of its component indicators. For more on return periods in general and this method in particular, see WSIM Concepts.

Figure 1 and Figure 2 present results contrasting the two compositing methods. Note that the adjusted composite tends to remove large areas of “abnormal” 3-to-5-year deficits and surpluses while preserving the more extreme anomalies. This also tends to reduce the extent of regions experiencing indications of both surplus and deficit.

Figure 1. WSIMv2 composite surplus and deficit indicators for the 12-month period December 2018 through November 2019 using the original WSIMv1 compositing method (left) and WSIMv2 using the adjusted composite (right). Maps are based on three months… — Figure 1. WSIMv2 composite surplus and deficit indicators for the 12-month period December 2018 through November 2019 using the original WSIMv1 compositing method (left) and WSIMv2 using the adjusted composite (right). Maps are based on three months of observed and nine months of forecast temperature and precipitation for the United States.

Figure 2. WSIMv2 composite surplus and deficit indicators for the 12-month period December 2018 through November 2019 using the original WSIMv1 compositing method (left) and WSIMv2 using the adjusted composite (right). Maps are based on three months… — Figure 2. WSIMv2 composite surplus and deficit indicators for the 12-month period December 2018 through November 2019 using the original WSIMv1 compositing method (left) and WSIMv2 using the adjusted composite (right). Maps are based on three months of observed and nine months of forecast temperature and precipitation for Europe.

DATA INPUTS
Some of the data sets and associated processing that are input to WSIM for production of the Global Water Monitor & Forecast Watch List reports have changed between WSIMv1 and WSIMv2. Table 1 lists WSIM data requirements and changes to the data sources made in WSIMv2. The most important difference is the use of a new source for soil water holding capacity.

Figure 3 compares WSIMv1 and WSIMv2 composite surplus and deficit “hot spot” maps of the United States for the same time, integration period, and the original composite method. The WSIMv1 and WSIMv2 maps are consistent with one another. However, there are modest differences due to the changes in input data. Figure 4 presents the comparable results for Europe.

Figure 3. Composite surplus and deficit indicators using the original composite method for the 12-month period December 2018 through November 2019 generated by WSIMv1 (left) and WSIMv2 (right). Maps are based on three months of observed and nine mon… — Figure 3. Composite surplus and deficit indicators using the original composite method for the 12-month period December 2018 through November 2019 generated by WSIMv1 (left) and WSIMv2 (right). Maps are based on three months of observed and nine months of forecasted temperature and precipitation for the United States.

Figure 4. Composite surplus and deficit indicators using the original composite method for the 12-month period December 2018 through November 2019 generated by WSIMv1 (left) and WSIMv2 (right). Maps are based on three months of observed and nine mon… — Figure 4. Composite surplus and deficit indicators using the original composite method for the 12-month period December 2018 through November 2019 generated by WSIMv1 (left) and WSIMv2 (right). Maps are based on three months of observed and nine months of forecasted temperature and precipitation for Europe.

DATA TABLE
Table 1. WSIM data requirements and data source changes between WSIMv1 and WSIMv2.

Data	Source WSIMv1	Source WSIMv2
Temperature (monthly, observed)	Global Historical Climatology Network (GHCN) + Climate Anomaly Monitoring System (CAMS) [1]	No change
Precipitation (monthly, observed), Number of wet days	NOAA’s PRECipitation REConstruction over Land (PREC/L)[2]	No change
Precipitation (daily, observed)	NOAA/CPC Unified Gauge-Based Analysis of Global Daily Precipitation[3]	No change
Precipitation (monthly, forecast)	NOAA’s Climate Forecast System Version 2 (CFSv2)[4]	No change
Temperature (monthly, forecast)	NOAA’s Climate Forecast System Version 2 (CFSv2)	No change
Soil water holding capacity	Harmonized World Soil Database v1.1[5] ISIRC WISE v3.1 soil profile database[6]	ISIRC WISE Derived Soil Properties[7]
Basin Delineation	Global Drainage Basin Database (GDBD)[8]	HydroBASINS[9]
Flow Direction Grids	ISLSCP II Simulated Topological Network (STN-30P)[10]	No change
Terrain Elevation	SRTM30 global enhanced elevation[11]	GMTED2010[12]

[1] Fan, Y. & van den Dool, H. (2008). A global monthly land surface air temperature analysis for 1948-present. Journal of Geophysical Research, 113. doi: 10.1029/2007JD008470.

[2] Chen, M., Xie, P., Janowiak, J.E. , & Arkin, P.A. (2002). Global Land Precipitation: A 50-yr monthly analysis based on gauge observations. Journal of Hydrometeorology, 3, 249-266.

[3] CPC Global Unified Precipitation data provided by the NOAA/OAR/ESRL PSD, Boulder, Colorado, USA, from their Web site at https://www.esrl.noaa.gov/psd/.

[4] Saha, S., Moorthi, S., Wu, X., Wang, J., Nadiga, S., Tripp, P., ... Becker, E. (2014). The NCEP climate forecast system version 2. Journal of Climate, 27, 2185-2208. doi: http://dx.doi.org/10.1175/JCLI_D-12000823.1

[5] FAO, IIASA, ISRIC, ISS-CAS, JRC. (2009). Harmonized World Soil Database (version 1.1). FAO: Rome, Italy and IIASA: Laxenburg, Austria.

[6] Batjes, N.H. (2008). ISRIC-WISE harmonized global soil profile dataset (Ver. 3.1). (2008). ISRIC – World Soil Information Center: Wageningen, The Netherlands.

[7] Batjes, N.H. (2016). Harmonized soil property values for broad-scale modelling (WISE30sec) with estimates of global soil carbon stocks. Geoderma, 269, 61-68. ( https://doi.org/10.1016/j.geoderma.2016.01.034

[8] Masutomi, Y., Inui, Y., Takahashi, K., & Matsuoka, Y. (2009). Development of highly accurate global polygonal drainage basin data. Hydrological Processes, 23, 572–584. doi:10.1002/hyp.7186. https://doi.org/10.1002/hyp.7186

[9] Lehner, B. & Grill, G. (2013). Global river hydrography and network routing: baseline data and new approaches to study the world’s large river systems. Hydrological Processes, 27(15), 2171–2186.

[10] Vörösmarty, C.J. & Fekete, B.M. (2011). ISLSCP II River Routing Data (STN-30p). In Hall, Forrest G., Collatz, G., Meeson, B., Los, S., Brown de Colstoun, E., & Landis, D. (eds.). ISLSCP Initiative II Collection. Data set. Available on-line [http://daac.ornl.gov/] from Oak Ridge National Laboratory Distributed Active Archive Center, Oak Ridge, Tennessee, U.S.A. doi.org/10.3334/ORNLDAAC/1005.

[11] ISciences, LLC. (2003). TerraViva! SRTM30 global enhanced: elevation, slope, aspect. ISciences, Ann Arbor, MI.

[12] Tyler, D.J., & Greenlee, S.K. (2012). Creation of digital contours that approach the characteristics of cartographic contours: U.S. Geological Survey Scientific Investigations Report 2012–5167.