Drought Monitoring Program

After the creation of the National Drought Mitigation Center (NDMC) in 1995, one of our first goals was to create a "one-stop shopping" section on our Web site that would provide users with access to all the information for drought monitoring in a timely and reliable fashion. The development of the "drought watch" section was undertaken because no routine national or regional integrated assessment was available from federal agencies. This section has evolved into the "Monitoring Drought" section of the NDMC Web site (drought.unl.edu). Currently, drought monitoring in the United States relies heavily on a product called the U.S. Drought Monitor, which draws on the climate and water supply indicators and indices such as Palmer drought severity index (PDSI) and the standardized precipitation index (SPI).

Palmer Drought Severity Index

The PDSI was developed by W.C. Palmer (1965) for monitoring droughts in terms of their intensity, duration, and spatial extent. Although there are several variations of the index, each variation has characteristics similar to the PDSI (Palmer, 1965; Karl and Knight, 1985; Heddinghaus and Sabol, 1991). Palmer based the PDSI on anomalies in the supply and demand concept of the water balance equation. Inputs into weekly or monthly calculations include precipitation, temperature, and the local antecedent soil moisture conditions. The data are standardized to account for regional differences so that the PDSI values can be compared from one location to another. Therefore, identical PDSI values, in theory, in the Midwest and Texas indicate the same severity of drought, even though actual rainfall deficiencies would be different at the two locations.

Weekly maps of a modified PDSI (Heddinghaus and Sabol, 1991) are produced by the Climate Prediction Center (CPC) of the National Oceanic and Atmospheric Administration (NOAA) and are frequently used in assessments of agricultural conditions around the United States. The PDSI has been used historically in policy decisions by the U.S. Department of Agriculture (USDA) regarding requests for drought relief and by states as triggers for response actions as part of state drought plans. However, the PDSI has limitations that diminish its application, bringing into question the practice of basing agricultural policy decisions solely on the PDSI. These limitations have been well documented (Alley, 1984; Karl and Knight, 1985; Willeke et al., 1994; Kogan, 1995; McKee et al., 1995; Guttman, 1998). The most significant limitations of the PDSI related to monitoring agricultural drought include: (1) an inherent time scale (about 8 months) that makes it difficult to detect emerging drought conditions and shorter-length drought periods; (2) the characteristic that all precipitation is treated as rain, which does not account for snowfall, snow cover, and frozen ground, thus questioning the reliability of the real-time winter PDSI values and preplant soil moisture estimates (one of the PDSI inputs); (3) the characteristic that the natural lag between precipitation and runoff is not considered and that no runoff occurs until the water capacities of the surface and subsurface soil layers are full, which leads to an underestimation of runoff; and (4) wide variations in the "extreme" and "severe" classifications of the PDSI values, depending on location in the country.

If a drought index is going to be spatially comparable and useful for agricultural policy decisions, extreme and severe classifications must occur consistently and with low frequency (Guttman et al., 1992). An additional concern is that the PDSI does not do well in the mountainous western United States, especially since a majority of that region's precipitation falls during the winter as snowfall. The PDSI can also retain values reflecting drought well after a climatological recovery from drought has occurred. All of these limitations reveal the importance of caution when using the PDSI for monitoring agricultural conditions and making related policy decisions.

Standardized Precipitation Index

The SPI was developed to address some of the problems inherent in the PDSI. McKee and his colleagues at Colorado State University (McKee et al., 1993, 1995) designed the SPI to be a relatively simple year-round index for monitoring drought and water supply conditions in Colorado. The SPI supplemented information provided by the PDSI and the surface water supply index (SWSI) developed by Shafer and Dezman (1982). The SPI is based on precipitation alone, whereas the SWSI incorporates snowpack, streamflow, precipitation, and reservoir storage. Calculation of the SPI for the specified time period for any location requires long-term monthly precipitation data for at least 30 years (i.e., the longer the data set, the more reliable the SPI values). The probability distribution function is determined from the long-term record by fitting a function to the data. The cumulative distribution is then transformed using equal probability to a normal distri bution with a mean of zero and standard deviation of one so the values of the SPI are expressed in standard deviations (Edwards and McKee, 1997). A particular precipitation total for a specified time period is then identified with a particular SPI value consistent with probability. Positive SPI values indicate greater than median precipitation, whereas negative values indicate less than median precipitation. The magnitude of departure from zero represents a probability of occurrence so that decisions can be made based on this SPI value. An SPI value of less than -1.0 (moderately dry) occurs 16 times in 100 years and an SPI of less than -2.0 (extremely dry) occurs 2-3 times in 100 years.

The fundamental strength of the SPI is that it can be calculated for a variety of time scales. This versatility allows the SPI to be used to monitor short-term water supplies such as soil moisture, which is important for agricultural production, and longer-term water resources such as ground water supplies, streamflow, and lake and reservoir levels, which are important for agriculture and other water users. Colorado uses the SPI information as part of a routine climatic assessment completed by the Water Availability Task Force for Colorado's drought plan. This information is useful for detecting the potential impacts of drought on agriculture and other economic sectors. Determining the linkages between SPI values at different time scales is the subject of considerable research as those involved in monitoring drought seek to identify triggers to initiate various mitigation actions for agriculture and other sectors.

The SPI has a number of advantages over the PDSI. First, it is a simple index and is based only on precipitation. The PDSI calculations are complex because 68 terms are defined as part of the calculation procedure (Soule, 1992). In spite of the complexity of the PDSI, McKee (personal communication, 1996) believes that the main driving force behind the PDSI is precipitation. Second, the SPI is versatile. It can be calculated on any time scale, which gives the SPI the capability to monitor conditions important for both agricultural and hydrological applications. This versatility is also critical for monitoring the temporal dynamics of a drought, including its onset and termination, which has typically been a difficult task for other indices. Third, because of the normal distribution of SPI values, the frequencies of extreme and severe drought classifications for any location and any time scale are consistent. McKee et al. (1993) suggest an SPI classification scale (table 9.1). Fourth, because it is based only on precipitation and not on estimated soil moisture conditions as is the case with PDSI, the SPI is just as effective during the winter months.

Although developed for use in Colorado, the SPI can be applied to any location with a data set of 30 years or longer. SPI maps for multiple time scales are routinely available on the NDMC Web site (http://drought.unl. edu) in the "drought watch" section and on the Web site of the Western Regional Climate Center (http://www.wrcc.dri.edu/spi/spi.html). The NDMC has disseminated SPI information and software at workshops and through direct e-mail contact with foreign governments, international or-

Table 9.1 Classification of drought categories for the standardized precipitation index (SPI) (according to McKee et al., 1993)

SPI values

Drought category

Time in category

0 to -0.99

Mild drought

34.1%

-1.00 to-1.49

Moderate drought

9.2%

1.50 to-1.99

Severe drought

4.4%

<-2.00

Extreme drought

2.3%

-50%

ganizations, and nongovernmental organizations. It is now being used in both operational and research modes in more than 50 countries and has been proven quite effective as part of a comprehensive, integrated early warning system.

The NDMC's experience has been that the SPI detects emerging drought conditions more quickly than the PDSI, a characteristic that is extremely critical in the timely implementation of mitigation and response actions by individuals and governments (Hayes et al., 1999). Many states are using the SPI as part of their efforts to monitor drought and trigger various drought-related mitigation and response actions.

Developing an effective drought monitoring system presents many unique challenges because of the slow onset nature of drought, its spatial extent and duration, and the requirement that multiple indicators and indices be used to properly characterize its severity and potential impacts. In addition to these challenges, the ineffectiveness of drought monitoring systems in the United States and elsewhere is also associated with inadequacies in the systems themselves. First, monitoring systems have often depended on an inadequate network of weather stations. Data from these stations may be reported infrequently (e.g., monthly) so that information is not readily available to decision makers at critical times or decision points. Second, drought monitoring systems are often based on a single parameter or index. Because of the complexities of drought, no single parameter or index can adequately capture the intensity and severity of drought and its potential impacts on a diverse group of users. Each index has strengths and weaknesses, which often vary spatially. Third, the delivery of information products to assess drought severity is often untimely. And fourth, information products are often developed without a clear understanding of user needs, or users are confused about how to apply this information when making critical climate-based decisions.

A comprehensive drought monitoring system has been recommended for many years in the United States (Wilhite et al., 1986; Riebsame et al., 1991; Wilhite and Wood, 1994), but no action on these recommendations had taken place until recently. In 1999 it became apparent that a new approach to drought monitoring was needed to address many of the inadequacies noted above.

U.S. Drought Monitor

The U.S. Drought Monitor was developed through a partnership between the National Drought Mitigation Center at University of Nebraska, the National Centers for Environmental Prediction/Climate Prediction Center (NCEP/CPC) of NOAA, and the USDA's Joint Agricultural Weather Facility (USDA/JAWF). The National Drought Policy Act, passed by the U.S. Congress in the summer of 1998, and the subsequent formation of the National Drought Policy Commission (NDPC) and its working groups in 1999 provided additional momentum to improve drought monitoring efforts in the United States. A working group on monitoring and prediction formed by the NDPC during spring 1999 provided additional opportunities for interactions with a larger group of climatologists throughout the country on drought monitoring issues. The group also helped to form the template for early versions of the U.S. Drought Monitor, first released on an experimental basis on May 20, 1999, as a biweekly product. The Drought Monitor became an operational product in August 1999 when it was officially released at a joint White House press conference conducted by the Department of Commerce and USDA.

The Drought Monitor is maintained on the Web site of the NDMC (http://drought.unl.edu/monitorLmonitor.html). It consists of a map showing which parts of the United States are suffering from various degrees of drought (figure 9.2). The map also accompanies text that describes the drought's current impacts, future threats, and prospects for improvement. The Drought Monitor is derived from several key parameters and ancillary indicators (e.g., fire potential, pasture and range conditions) from different agencies. The six key parameters making up the scheme at this writing are the PDSI, the Climate Prediction Center's Soil Moisture Model (per-centiles), the U.S. Geological Survey's daily streamflow (percentiles), percentage of normal precipitation, SPI, and a remotely sensed satellite vegetation health index. Table 9.2 illustrates the drought severity classification system currently used to prepare the map.

The authors of the Drought Monitor rely on different agencies for the inputs required to create the map. The initial draft of the map is produced on Monday and distributed via e-mail to more than 150 climate, water supply, and agricultural specialists throughout the country. These persons are asked to review the map and provide comments. These regional experts often have a better understanding of local situations because of their direct contacts with agricultural, water, and natural resources managers. Based on the comments from these reviewers, the map is revised and a second draft is distributed. The final map is completed by Wednesday night and placed on the Web site at 0730 h (Central Standard Time) each Thursday morning. Previous maps are archived, and users can also see an animation of the past 6- and 12-week periods to better visualize the changing spatial extent and severity of drought conditions across the country. The Drought Monitor map for May 21, 2002, provides an example (figure 9.2).

U.S. Drought Monitor »J1¿T2002

Figure 9.2 U.S. Drought Monitor for May 21, 2002.

The Drought Monitor map classifies droughts on a scale from one to four (D1-D4; i.e., from least to most intense droughts). D4 is a 1-in-50-year event. A fifth category, D0, indicates an abnormally dry area either heading into drought or recovering from it. The Drought Monitor also shows which sectors are experiencing the dominant or primary impacts, using the labels A (agriculture: crops, livestock, range, or pasture), W (water supplies), or F (high risks of fire danger), as shown in figure 9.2. For example, an area shaded and labeled D2 (A) is in general experiencing severe drought conditions that are affecting the agricultural sector more significantly than the water supply sector. The area is not seeing a heightened fire risk in association with this dryness. An area shaded as D2 with no A, W, or F would be experiencing impacts in all three sectors. The final map summarizes all of this information in an easy-to-read format that shows where drought is emerging, lingering, or subsiding.

This map product has been widely accepted and is used by a diverse set of users to track drought conditions across the country. The users include agricultural producers; commodity brokers; water and natural resource managers; congressional delegations; local, state, and federal agencies; and mainstream media. The number of hits on the Drought Monitor Web site increased from 1.75 million in 2000 to more than 5 million in 2002, exceeding our greatest expectations. The Drought Monitor has already become an essential component of the United States's

Figure 9.2 U.S. Drought Monitor for May 21, 2002.

Released Thursday, May 23, 2002 Autlioil Mir* Syobodl, NDMC

Table 9.2 Drought Monitor's drought severity classification system

Drought type Associated ranges of objective indicators

Table 9.2 Drought Monitor's drought severity classification system

Drought type Associated ranges of objective indicators

CPC soil

USGS

Palmer

moisture

weekly

Percent

Standardized

Satellite

drought

model

streamflow

of normal

precipitation

vegetation

Category

Description

index

(percentiles)

(percentiles)

precipitation

index

health index

DO

Abnormally dry

-1.0 to-1.9

21-30

21-30

<75% for 3 months

-0.5 to -0.7

36-45

D1

Moderate drought

-2.0 to-2.9

11-20

11-20

<70% for 3 months

-0.8 to-1.2

26-35

D2

Severe drought

-3.0 to-3.9

6-10

6-10

<65% for 6 months

-1.3 to-1.5

16-25

D3

Extreme drought

-4.0 to -4.9

3-5

3-5

<60% for 6 months

-1.6 to-1.9

6-15

D4

Exceptional drought

-5.0 or less

0-2

0-2

<65% for 12 months

-2.0 or less

1—5

initiative toward a national drought policy (Wilhite, 2001; Svoboda et al., 2002). For example, the states significantly affected by drought in 2000, 2001, and 2002 used the product to track the severity of drought conditions in their state and make policy decisions on emergency and mitigation actions. Users appreciate a product that simplifies a difficult and complex issue but is based on scientific climatic indices and parameters.

Certainly, the Drought Monitor cannot always capture the local situation accurately. The partners in this activity are striving to improve the science of drought monitoring by improving networks and developing new climatic indices and other assessment tools to make this a better product in the future.

0 0

Post a comment