The Combined Drought Indicator (CDI) system, as detailed and outlined by Sepulcre-Canto et al. in their comprehensive 2012 study, presents a unique and insightful progression of the phenomenon known as agricultural drought through a clear cause-effect relationship. This model, with its basis on critical empirical data, makes the inference that a significant lack of precipitation, or rain, inherently leads to a substantial deficit in soil moisture.

This deficit, in turn, has a direct and profound impact by lowering the productivity of vegetation, a crucial aspect of agricultural processes. The CDI system, through its intricate and well-defined methodologies, describes the various stages and phases of an agricultural drought. It does this by employing and analysing three primary variables, each playing a pivotal role in the process.

These variables include: the Standardised Precipitation Index (zSPI), which measures the deviation in precipitation; soil moisture (zSM), an indicator of the water content in the soil; and the fraction of absorbed photosynthetically active radiation (zfAPAR), a key factor in the process of photosynthesis and plant growth.

The stages of the CDI are as follows:

Strengths: The spatial coverage is good and at a high resolution using a combination of remotely sensed and surface data.

Weaknesses: Using a single SPI value may not be the best option in all situations and does not represent conditions that may carry over from season to season

Housed and maintained at the European Drought Observatory within the European

Commission’s Joint Research Centre, http://edo.jrc.ec.europa.eu/edov2/php/index.php?id=1000.

Sepulcre-Canto, G., S. Horion, A. Singleton, H. Carrao and J. Vogt, 2012:

Development of a Combined Drought Indicator to detect agricultural drought in Europe. Natural Hazards and Earth Systems Sciences, 12:3519–3531.

We generate monthly precipitation maps at a spatial resolution of 5566 m using data from the CHIRPS Pentad: Climate Hazards Group InfraRed Precipitation With Station Data. This dataset merges satellite imagery with in-situ station data to provide detailed precipitation estimates.

The formula for calculating the Standardised Precipitation Index (SPI) is based on well-established methods described in the literature. The process typically involves fitting a Gamma distribution to precipitation data, adjusting for the probability of zero precipitation, and transforming the cumulative distribution function (CDF) into a standard normal distribution (z-score) using the inverse error function. Here are the key references that describe this methodology:

​

This process transforms precipitation data into a normalised index (SPI) that can be used to assess wet and dry periods consistently across different climatic regions.

The 1-month and 3-month Standardised Precipitation Indices (SPI-1 and SPI-3) are used to identify drought conditions. These indices utilise the two-parameter gamma distribution fitted over a 30-year reference period (1981–2010) and calculated using maximum likelihood estimators (Thom, 1958; Greenwood and Durand, 1960). The SPI-3 is particularly relevant for vegetation drought monitoring because of its relationship with drought conditions (WMO, 2012), whereas the SPI-1 is handy for detecting flash droughts when high temperatures, low humidity, or strong winds enhance evaporative demand (Otkin et al., 2018).

We set threshold values at -1.0 for SPI-3, marking the beginning of moderately dry conditions (McKee et al., 1993), and at -2.0 for SPI-1, indicating the start of extremely dry conditions. For simplicity, we define a Boolean SPI indicator (zSPI) in the CDI framework, which takes a value of 1 if either SPI-1 or SPI-3 signify dry conditions. This comprehensive approach ensures that the CDI system effectively captures both short-term and medium-term precipitation deficits, improving the accuracy of drought detection and monitoring.

The soil moisture anomaly index (zSM) is computed using soil moisture data from the SPL4SMGP.007 SMAP L4 Global 3-hourly 9-km Surface and Root Zone Soil Moisture dataset. Initially, decadal (approximately 10-day) maps of the soil moisture index (SMI) are created at a 11 km spatial resolution. These maps are computed as a weighted average of daily volumetric soil moisture values from the SMAP dataset for both the skin and root zone layers. Subsequently, the zSM is derived as standardised deviations (z-scores) of these values.

In this study, SMI is used to enhance map readability for non-expert users, as SMI ranges simply from 0 (dry) to 1 (wet). The zSM maps effectively capture deviations from the norm, allowing for a clear identification of drought conditions. A threshold of -1 is used to identify dry conditions in zSM, consistent with the threshold for SPI-3.

The SPL4SMGP.007 SMAP dataset, provided by NASA and the National Snow and Ice Data Center (NSIDC), includes several relevant parameters used in the calculations:

These parameters are crucial for calculating the SMI and subsequently the zSM, ensuring accurate monitoring and assessment of soil moisture conditions.

The fraction of absorbed photosynthetically active radiation (fAPAR) is a critical biophysical variable that measures the proportion of available light in the photosynthetically active spectrum (400-700 nm) absorbed by vegetation. This indicator is crucial for understanding plant health and productivity because it directly correlates with the process of photosynthesis, where plants convert light energy into chemical energy.

This study uses the fraction of absorbed photosynthetically active radiation (fAPAR) as an indicator for vegetation health, estimated from satellite remote sensing data. The MOD15A2H Collection 6 fAPAR product from the Moderate-Resolution Imaging Spectroradiometer (MODIS) sensor on the Terra satellite is implemented in the Combined Drought Indicator (CDI).

NASA provides the MOD15A2H product. It has a spatial resolution of 500 meters and comes as 8-day maximum composites. The European Drought Observatory (EDO) re-projects these raw data onto a 0.01° latitude and longitude grid. Decadal maps are made using a weighted average of the two closest 8-day maps and then applying exponential smoothing. fAPAR anomalies (zfAPAR) are calculated as standardised z-scores from the complete dataset spanning 2016 to 2024. A threshold value of -1.0 signifies dry conditions.

​

Drought risk areas - green (low), yellow (medium), red (high)