Projects App

DSDI: stands for Demand Sensitive Drought Index. Refer to the INTRO panel for a detailed description!

Cumulative Deficit: The deficit (Demand-Supply) cumulated over the entire period of analysis [cumulated when positive; set to zero when negative].

Drought Event: A part of the cumulative deficit curve above the zero line constitutes a drought event.

Threshold Concept: While in many drought studies they adopt a threshold to define a drought event, we do not consider the threshold approach to define drought events in this research. The zero line is the basis to define a drought event. However, the approach can be implemented by considering the reserve at the location (total water store) as baseline.

Drought Attributes: We define four attributes for a drought event: Onset, Duration, Severity, and Recovery. The onset of the drought event is the point in time where the deficit begins (greater than zero). The Duration is the period from the onset to the maximum deficit; the maximum deficit is the severity; and the period from the maximum deficit until the cumulated deficit drop back to zero is the recovery.

Worst Drought: The worst drought can be defined in terms of its Duration, its severity, or its Recovery time. It is the drought event with the highest value of the considered attributes.

NDC, NDImax: The Normalized Deficit Cumulated (NDC) index capture droughts resulting from deficits carried over years (multi-year droughts), and athe NDImax capture within year droughts. There are both computed as the ratio of the maximum deficit to the average precipitation received by the location over the period of analysis. NDC<1 indicates that the location receives on average enough rainfall to cover its deficit. NDC>1 means the amount of rainfall is not sufficient. The location is then dependent on other water sources to meet the demand.

Resiliency: the probability of recovery from failure is termed as Resiliency (or the recovery rate). It indicates how often drought events are fully recovered. It is computed as the ratio of the number of drought events to the total duration time of all events. A county with a higher Recovery Rate compare to another has a greater resiliency to drought. Droughts in this county recovers often and do not last for long.

Relative Recovery: We define the relative recovery as the expected value of the ratio of the drought recovery time to the drought duration time. If the recovery rate < 1 drought events recover quicker than they last (in average).

The app display a map and summary statistics (using boxplots) of droughts metrics and attributes computed at the county level for the State of California.

The daily water deficit is estimated (and accumulated) using daily water supply over 101 years (from 1910 to 2010) and water withdrawal (demand), which is assumed constant for the whole period of analysis. The water withdrawal is obtained from the USGS water use database (provide link). The data is collected every 5 years starting from 1985.  We assumed the water withdrawal every 5 year constant over the period of analysis to study the impact of each demand scenario, put in the context of California's climate variation. The app display a map the at the county scale for the selected values on the right panel.

Default Display [refresh page if necessary]: The default selections display a map of the Resiliency Rate of Drought computed using Total water usage for the year 2010 as the constant water demand scenario. Three groups of boxplots show respectively from left to right, the metric selected (NDC), the demand, and the drought attribute selected (Duration) for two demand scenarios (1985, and 1990) . This allows comparison across different demand patterns (scenarios). For instance, the boxplot showing the duration indicates that the worst drought has shorter duration times using the 2010 demand scenario in comparison to the 1985 demand scenario.

The map shows the spatial distribution at the county level of the selected 'Drought metric' for the 'Chosen Crop'. The checkboxes labeled 'Compare Crops' allow the comparison of crops for the selected drought metric or for one of the 'Drought Attributes' shown at the bottom of the right panel. Two group of boxplots show the comparison results across the selected crops. Each group of boxplots shows respectively the comparison for the drought metric and the drought attribute selected.

Default display: The map shows, at the county level, the maximum accumulated water deficit (normalized by the annual average rainfall for each county) over the period from 1949 to 2009 (computation at the daily scale). It is showing that most counties in the mid-west United States (orange and red) would be experiencing water stress if the agriculture over the region was rain-fed. In average, the annual rainfall is not enough to cover the demand. For those counties the use of groundwater or other sources of water (such as water transfers), to compliment the endogenous water of the county, is evident. The boxplots on the left compare the value of NDC for Corn and Soybeans. Higher values of NDC are observed for Corn than for Soybeans. On the right, the boxplots compare the drought duration for the two crops. We can observed that the drought duration for Corn is higher than for Soybeans, and it is sometimes the maximum number of years (62) over the period analyzed. The drought do not recover in the case of those counties.

The other drought metrics:

  • NDImax: NDImax indicates the maximum intra-annual accumulated water deficit (normalized by the annual average rainfall). The deficit is accumulated for each year instead of for the whole period as for NDC. The deficit is then not carried over the years but considered to be zero at the beginning of each year.
  • Resiliency rate: This index is the probability of recovery for a county. It is estimated based on the daily accumulated deficit time series [see paper for details]. Counties more probable to recover from droughts (typically counties with bursting rainfall events) are more resilient [the higher the value, the better].
  • Relative Recovery: This indicates whether drought events recover faster than they last on average [the lower the value, the better].

(*) climate: i.e. the evaporative demand of the atmosphere for the crop

In this section, DSDI is computed using  the evaporative demand of the atmosphere for the selected crop (as opposed to the crop's actual demand as computed in the 'DSDI CONUS' tab).

The top map shows the response of the crop based on both fluxes of precipitation and the climate demand. The green and yellow color patterns indicate where the stress is less for growing the specific crop (under rain-fed conditions). Where the stress is higher (orange and red patterns), irrigation is needed to complement the precipitation. The map can be used to identify locations that are more suitable to grow a specific crop in terms of water requirements.

The bottom map shows where the crop is actually planted (based on the USDA NASS survey). Notice that the crop is sometimes planted in high stress locations.