Drought monitoring increasingly takes place on smaller and smaller scales, intensifying the demand for localized information. To help meet this growing need, McRoberts and Nielsen-Gammon (2012) developed a computational procedure for calculating the Standardized Precipitation Index (SPI) that incorporates high-resolution, radar-based estimates of precipitation. The SPI is a precipitation-based drought index that relates the amount of precipitation falling over a given interval of time to its historical probability. SPI has several key advantages for drought monitoring, namely its ability to be calculated on any timescale, provided sufficient data is available, giving it the ability to monitor simultaneously-occurring conditions at different time scales. Additionally, SPI values are normalized to the historical record at a specific location, meaning both wet and dry periods can be monitored and SPI values at different locations can be compared without modification.
The SPI calculation methodology developed by McRoberts and Nielsen-Gammon (2012) incorporates daily gridded precipitation estimates from the National Weather Service’s (NWS’s) Advanced Hydrologic Prediction Service (AHPS), monthly precipitation normals from the Parameter-elevation Regression on Independent Slopes Model (PRISM), and historical gauge data from the NWS Cooperative Observer Program (COOP) network. McRoberts and Nielsen-Gammon (2012) developed and evaluated this SPI calculation methodology over Texas. In partnership with Texas A&M University and Purdue University, NC State University was granted funding to expand this work to more regions, evaluate its performance, and develop a web-based interface for users to access and interact with the resulting SPI grids. Development of the SPI and initial web interface was supported by USDA NIFA and NIDIS. Ongoing support for development of SPEI and Palmer calculations is provided by NOAA CPO.
Historical distribution parameters were determined using historical accumulated precipitation time series from NWS COOP stations from across the contiguous US. Stations from this network are chosen because of the long term records of many sites as well as for consistency with methodology established by McRoberts and Nielsen-Gammon (2012). COOP stations having at least 60-year record lengths and a minimum of 90% of records over the 1981-2010 normals period were obtained from the Climate Retrieval and Observations Network of the Southeast (CRONOS) database, which contains historical and current observations from COOP stations across the US using the NCEI’s TD3200 data (NCAR 1981).
Following initial selection, COOP stations are clustered using Ward’s clustering technique with subjective adjustments (Hosking and Wallis 1997). Regional frequency analysis was employed to calculate the parameters of the historical Pearson Type III distribution at each station and cluster using L-moment ratios. For this, each station’s period of record, (or since 1 January 1895 for stations with longer records) was used. Distribution parameters were calculated using L-moment ratios. Following calculation, they were normalized by dividing by the location parameter, and 1981-2010 PRISM normals are used as the historical mean of the distribution.
Once distribution parameters were determined for each site, they were interpolated to the HRAP grid, the same grid that AHPS precipitation is available on.
For a given date, timescale, and cluster solution, AHPS Precipitation is summed for all the dates during the time period, then divided by the corresponding PRISM normal precipitation to calculate the fraction of normal. This is combined with the normalized distribution parameters for the given timescale and ending calendar day to determine the cumulative probability of the observed precipitation amount. Using guidance from Hosking (2005), the inverse normal (Gaussian) function is then applied to calculate the SPI.AHPS Precipitation
AHPS Precipitation estimates (water.weather.gov/precip) are used to generate gridded precipitation estimates. These grids have data over the contiguous US for 24-hour periods ending at 12Z (7am EST) each day. Additionally, AHPS precipitation grids, the SPI generated using them, are available on the HRAP grid, which is defined on a polar stereographic map projection which has a grid length of 4.7625 km along the standard parallel of 60◦N (Reed and Maidment 1999). NWS River Forecast Centers (RFCs) generate gridded quantitative precipitation estimates (QPE) over their domain and these are mosaicked into the national gridded precipitation products.
Each RFC uses a slightly different approach to quality control the data over its domain, and the approaches used to generate gridded precipitation estimates at the RFCs can be broadly classified into two categories: the three westernmost RFCs (NWRFC, CNRFC, and CBRFC) use the PRISM/Mountain Mapper approach while the 9 eastern RFCs use a multisensor precipitation estimate (MPE) approach (AHPS/NWS 2015). The PRISM/Mountain Mapper approach combines hourly surface precipitation gauge observations with monthly PRISM normals to determine gridded estimates of hourly precipitation. MPE is a combination of radar-derived precipitation, surface gauge measurements, and occasionally, satellite estimates of precipitation (Lin and Mitchell 2005). The main advantage of the high-resolution SPI generated here is the incorporation of radar-estimated precipitation, which has the ability to capture precipitation in locations where no or few surface gauges are located.
AHPS Precipitation is available since 2005. Since AHPS precipitation grids are generated daily in real-time, the SPI calculated using them can also be updated daily and generated in an operational environment useful for drought monitoring.PRISM Monthly Normals
PRISM (Parameter-elevation regressions on independent slopes model) grids from Oregon State University of 1981-2010 monthly normals data (Daly et al. 2008; PRISM Climate Group 2013) are used as the location parameter of the historical probability distribution function for generating the gridded SPI discussed here. PRISM assumes that elevation is the most important factor in the spatial distribution of precipitation (Daly et al. 1994, Daly et al. 2002). These data are available on latitude-longitude geographic coordinates, and are reprojected to the HRAP grid to allow for simple integration with the gridded precipitation data. To determine the PRISM normals that cover moving-window durations spanning across month(s), the monthly PRISM normals are divided by the number of days in the given month and these are then multiplied by the number of days of each month needed to fill the period.
An experimental SPI calculation, known as SPI Blend, assumes that more recent precipitation has a great influence on the current severity of dryness (or wetness) than precipitation that fell farther in the past. The SPI Blend accounts for this by giving greater weight to more recent precipitation in the calculation of historical distributions as well as observed precipitation amounts.SPEI (beta)
Note: These grids are still being updated and evaluated; they should be used on an experimental basis only. The Standardized Precipitation Evapotranspiration Index is the newest addition to the HiRDTT. Computationally similar to SPI, the SPEI is based on the climate balance between precipitation (P) and potential evapotranspiration (PET). Like the SPI, SPEI values represent standard deviations above or below the historical P-PET value for the given timescale. Because SPEI incorporates a temperature component, it can identify droughts that are exacerbated by heat. SPEI is updated daily for time scales ranging from 30-days to 36-months. Daily PET is estimated using daily temperatures from PRISM. Occassionally there are delays in the availability of this data. In these instances, daily RTMA temperature data is utilized until the daily PRISM estimates become available.KBDI
The Keetch-Byram Drought Index is the only drought index with a fixed timescale. KBDI is typically used by foresters to assess the climatological potential for fire. KBDI values are typically lowest in the winter and highest in the summer. Values are based on the previous day's value and are adjusted up or down depending on the current day's temperature and rainfall amounts. KBDI has no timescale, so adjusting the dropdown menu for timescale will not adjust the KBDI map displayed.Percent of Normal Precipitation
This is the amount of precipitation over a given timescale divided by the normal amount for the same period, then multiplied by 100 to yield a percent. Daily AHPS precipitation estimates are used to calculate the observed precipitation amount. Monthly 1981-2010 PRISM normals are divided by the number of days in a month, then multiplied by the number of days needed to fill each period to calculate moving-window normal precipitation amounts. Percent of Normal Precipitation is updated daily for every available timescale from 1 day to 36 months.
A website was developed for viewing and interacting with gridded drought index data. New grids are generated daily and are typically available by 12pm EST. Having new grids each day allows climatologists, meteorologists, water managers, and others to monitor conditions as they evolve. The sub-county detail provided by the product yields valuable information about local conditions, and helps identify areas that may be experiencing degradation or improvements.
When users first come to the webpage, they are presented with a map displaying the SPI, with information about the date, timescale, and legend at the bottom of the map. On the left is a menu where users can change the location, date, and timescale. Clicking ‘Submit Options’ will redraw the map with the new selections.
Beneath this menu is a layer switcher where users can choose between displaying four different gridded layers: SPI, SPI Blend, SPEI (still in development), Percent of Normal Precipitation, Accumulated Precipitation, and Keetch-Byram Drought Index (KBDI). Additional boundary layers as well as the US Drought Monitor map and accumulated surface gauge precipitation from NWS COOP stations can be overlaid. This gives users the ability to cross-reference the gridded products with the current USDM depictions and surface observations, allowing users to make a more informed determination about the current severity and extent of dryness. All these options can be “saved” by clicking the “Bookmarkable Link” link on the lower left of the page. A document with all the possible URL inputs has also been compiled.
Users wishing to have more interaction with the SPI data can click the “Download GeoTIFF” link in the lower right corner of the map webpage to download a GeoTIFF, in geographic coordinates, of the map currently being viewed. This can easily be uploaded into ArcMap, Quantum GIS, or other GIS software.
In addition to viewing maps for their desired location, users can click on the map to place a marker and query for the SPI (or other gridded layer) value at that particular grid point. This will also produce a link that will take users to a page where the time series of SPI for the particular grid point can be viewed.
On the times series page, users can mouse-over the graph to see values and click-and-drag to zoom in. Links to the lower right of the graph can be used to download the time series data as a comma-separated list or image as a PNG. Like the map page, the time series page has various options that can selected and even bookmarked.
More advanced users can access the SPI grids, which are stored as NetCDF files, on a THREDDS server housed by the State Climate Office of North Carolina. Grids can be accessed via HTTPS or NetCDF subset service (for downloading), WMS and WCS for viewing in web- or GIS-based applications, or query grids using OPeNDAP for Ascii or binary data.
Development of the high-resolution drought index grids over the contiguous US was supported by USDA NIFA, NIDIS, and NOAA CPO. Additional input on the website as well as support to generate gridded Palmer Drought Indices has been provided by the Carolinas Integrated Sciences and Assessments (CISA), a NOAA RISA program. We would like to acknowledge the National Weather Service and the PRISM Climate Group for openly providing the datasets used in the calculations of the drought indices found within this tool.