About Us Advertise Careers Contact Us Privacy Policy Community Guidelines
Locating… 13°C English WhatsApp Alerts Email Alerts
BREAKING NEWS
Kieron Pollard Emulates Jos Buttler's World Record; Becomes First Player In History To Claim Rare Feat Why travellers are flocking to the world’s hardest-to-reach spots Sri Lanka's organised crime bill gets Supreme Court nod - The Hindu Live coverage: SpaceX to launch intelligence-gathering satellites for the National Reconnaissance Office Sri Lanka’s digital ID ambitions face a bigger test: Making government data work - Biometric Update Taming blue light: Researchers bring precision lasers onto a chip Has AI's honeymoon ended on Wall Street? Pakistani youth rank among world's most environmentally optimistic, ahead of India, BD Can teams of AI agents recreate the chain reaction of human decisions? How not to be threatened by memoirs
SRI LANKA NEWS

Drought pattern under climate change in Harris County, Texas, USA based on CMIP6 projections - Nature

Lanka Hit News • Mar 25, 2026 · 8mo ago • 43 min read
Share:

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Scientific Reports volume 16 , Article number: 14948 ( 2026 ) Cite this article

This study assessed future drought conditions in Harris County, Texas, using five Coupled Model Intercomparison Project Phase 6 (CMIP6) General Circulation Models (GCMs) under SSP1-2.6, SSP2-4.5, and SSP5-8.5 scenarios for the period 2026–2085. The analysis employed the Standardized Precipitation Evapotranspiration Index (SPEI-6 and SPEI-12) to characterize drought, providing comprehensive insights into future drought risk. The results indicate a substantial increase in drought frequency, intensity, and severity in Harris County, particularly during the far-future period (2056–2085). The study identified an oscillating drought pattern: high frequency (1980–2002), temporary reduction (2003–2024), and projected intensification (2056–2085). Under medium and high emission scenarios, both SPEI-6 and SPEI-12 exhibit significant ( p < 0.05) negative trends across the ensemble mean and most individual GCMs for the future period, in contrast to the historical period where both SPEI-6 and SPEI-12 demonstrate slight positive trends. Severity duration frequency curve analysis reveals that future droughts under SPEI-12 will be significantly more severe across all return periods (2–50 years). Although Harris County is experiencing relatively lower drought risk at present, the identified oscillating pattern and projected doubling of drought frequency under SSP5-8.5 represent a critical threshold for regional water security, requiring heightened attention from policymakers for proactive planning and adaptation strategies.

Climate change is intensifying drought frequency and severity globally, posing significant threats to water security and agriculture. Drought represents a naturally occurring phenomenon across virtually all climatic regions, identified by long periods of below-normal precipitation that constitute an inherent component of climate variability 1 , 2 , 3 . While drought has historically been a recurring natural hazard, climate change is intensifying its characteristics, leading to increased frequency, severity, and spatial extent 4 , 5 . Rising temperatures associated with climate change enhance potential evapotranspiration rates and amplify atmospheric water demand, creating conditions conducive to more intense and persistent drought episodes 6 . The global implications of this trend are already evident. Under high-emission projections, specifically the Shared Socio-economic Pathway scenario SSP5-8.5, global arid regions are projected to expand by roughly 10% by 2100 as a result of continued warming 7 . The worsening of drought conditions driven by climate change presents significant risks across economic, social, and environmental fields. These consequences are particularly evident in the disruption of economic productivity and community well-being, with agricultural sectors and rural economies bearing disproportionate impacts.

Texas exemplifies the challenges posed by intensifying drought conditions in a changing climate. The state’s climate is characterized by substantial variability and recurrent drought cycles, with notable historical drought periods including 2010–2011 8 . Compounding these natural patterns, temperatures across Texas have increased by 0.6 to 0.8 °C over the past century, contributing to more frequent and severe natural disasters 8 , 9 . While some regions of Texas, including the Houston metropolitan area, have experienced increased extreme precipitation events in recent years 10 , the state remains highly vulnerable to intensifying drought conditions under future climate scenarios. A study of Texas crops from 2008 to 2016 found that drought considerably reduces crop yields 11 , and it affects socioeconomic conditions.

General Circulation Models (GCMs) provide essential tools for evaluating future climate characteristics. Analysis of CMIP5 GCM outputs 12 demonstrated projected declines in surface water availability across the southwestern United States for 2021–2040, examining seasonal patterns in precipitation, evaporation, soil moisture, and runoff. This work was corroborated by 13 , who used identical CMIP5 data to project unprecedented 21st-century drought conditions for the region. Subsequently 14 , in their publication on unprecedented Drought Challenges for Texas Water Resources in a Changing Climate, presented a drying trend in the latter half of the 21 st century. The advancement from CMIP5 to CMIP6 models represents a significant improvement in climate projection capabilities, offering enhanced spatial resolution, improved representation of physical processes, and updated emissions scenarios 15 . Li et al. 16 , assessed future water availability across Texas using CMIP6 climate model data and found that the high-emission SSP5-8.5 scenario projects significant water availability declines across most of the state. Their analysis revealed severe net water deficits in southern Texas under both SSP2-4.5 and SSP5-8.5 scenarios, resulting from decreased precipitation and enhanced actual evapotranspiration, with the most severe conditions occurring along the border areas. However, these statewide assessments, while valuable, lack the spatial resolution and local context necessary for county-level water resource planning and adaptation strategies.

Harris County, encompassing Houston and ranking as Texas’s largest county 17 , represents a critical case for drought assessment due to its economic significance, rapid urbanization 18 , and exposure to both drought and flood extremes. Despite the clear vulnerability of Harris County to drought impacts and the critical importance of the region for economy and population, comprehensive assessments of future drought conditions at the county scale remain limited.

While Li et al. 10 , examined precipitation extremes in the Houston area using CMIP5 RCP scenarios, according to the literature, no study has yet assessed future drought conditions for Harris County using the latest CMIP6 SSP scenarios. This study addresses this gap by providing the first comprehensive drought assessment for Harris County using CMIP6 climate projections and multi-scale SPEI analysis.

Drought indices serve as indispensable tools for monitoring and assessing drought conditions. The Standardized Precipitation Evapotranspiration Index (SPEI) is widely used to identify drought events and is recommended by the IPCC AR6 for assessing agricultural and ecological droughts 19 , 20 . The SPEI is derived by fitting a log-logistic probability distribution to the cumulative water balance, defined as precipitation minus potential evapotranspiration (P - PET), across multiple temporal scales 21 . Comparative studies have shown that SPEI outperforms other drought indices, including the Palmer Drought Severity Index (PDSI), Palmer’s Z-index, precipitation percent normal, and precipitation percentiles, especially in capturing soil moisture conditions in the south-central United States 22 .

This study fills a critical research gap by utilizing recent CMIP6 climate model projections to evaluate potential future drought conditions in Harris County, Texas, across multiple time periods and emission scenarios. The research utilizes SPEI calculations across multiple temporal scales to characterize drought trend, frequency, severity, and intensity patterns under different climate change scenarios. The objectives of this study are to evaluate future drought projections under multiple CMIP6 scenarios and time periods, and quantify changes in drought characteristics, including frequency, severity, and intensity.

Geographical location and elevation variation of Harris County, Texas, showing NEX-GDDP-CMIP6 grid points used in the study. (The figure was generated using Arc GIS 10.4 using base maps from TIGER - https://catalog.data.gov/dataset/tiger-line-shapefile-2021-county-harris-county-tx , open source maps).

Harris County is located in the upper Gulf Coast in southern Texas, United States (Fig. 1 ). The City of Houston is housed in Harris County, which is the largest city in Texas 23 . The county spans a total area of 4600 km 2 , where 96% is land and 4% is water. The urban and built-up areas are the major land cover and pasture, woody savannas, cropland/natural vegetation mosaics, mixed forests are also major land use and land covers 24 . It is a county with dense urban areas and a high resident population per square mile compared to neighbouring counties ( https://datausa.io/profile/geo/harris-county-tx ). The local climate can be characterized as humid subtropical and springtime supercell thunderstorms, hot, humid summers, and temperate winters 17 . Harris County receives an average annual precipitation of approximately 1,200 mm 17 , 25 , while annual reference evapotranspiration ranges from 1,366 mm in the southeastern part to 1,500 mm in the western part of the county 24 , indicating that evaporative demand approaches or exceeds precipitation across the region. The average annual daytime land surface temperature is 30.01 °C, and the average annual nighttime land surface temperature is 17.83 °C 26 . The urban areas of Harris County continuously increase the demand for freshwater at an elevated rate, particularly in the northwestern regions 24 . In Harris County, surface water resources include major river systems such as the San Jacinto River, with its East and West Forks feeding Lake Houston and the Buffalo Bayou, which flows through central Houston toward Galveston Bay. Several tributaries and creeks such as Brays Bayou, Spring Creek, Clear Creek, Vince Bayou, and Cypress Creek further contribute to the river network and local drainage basins 27 .

Climate data from the selected GCMs were extracted from the NASA Earth Exchange (NEX) Global Daily Downscaled Projections (GDDP) dataset (NEX-GDDP-CMIP6) using ten grid points (P1 to P10) covering Harris County, as illustrated in Fig. 1 . Given the spatial resolution of 0.25° of the dataset, these ten grid points provided comprehensive coverage of the entire county area.

For the historical baseline scenario, daily precipitation, maximum, and minimum temperature data were obtained from the Daymet web portal 28 spanning the period from 1980 to 2024. These data served as the reference period for comparative analysis with future climate projections. The use of Daymet is supported by its extensive application in Texas hydroclimatic studies 29 – 32 .

For future climate projections in this drought assessment, we utilized the NEX-GDDP-CMIP6 data [ https://doi.org/10.7917/OFSG3345 ] 33 . This dataset provides bias-corrected and spatially downscaled climate data at a resolution of 0.25° × 0.25° (approximately 25 km × 25 km), processed using the Bias-Correction Spatial Disaggregation (BCSD) method 33 , 34 . The BCSD approach combines statistical downscaling with bias correction to remove systematic errors in GCM outputs while downscaling from coarse native resolution to finer spatial scales suitable for regional impact assessments. The reliability and accuracy of the NEX-GDDP dataset have been validated in previous studies 35 , 36 , 37 . For this drought assessment in Harris County, three Shared Socioeconomic Pathways (SSPs) were selected to represent distinct radiative forcing levels and socioeconomic development trajectories: SSP1-2.6 (sustainability pathway with low forcing), SSP2-4.5 (middle-of-the-road pathway with moderate forcing), and SSP5-8.5 (fossil-fueled development pathway with high forcing). Together, these scenarios capture a broad spectrum of potential future climate conditions for evaluating drought vulnerability under varying emission trajectories.

Through comprehensive statistical analysis, Yang 38 identified 15 top GCMs for Texas applications, while 39 detected several GCMs best suited for south central Texas. Additionally 16 , employed top-performing data-driven GCMs for Texas based studies. Abeysingha et al. 29 also utilised four GCMs for future climate change assessment in a watershed within Harris County. Deriving from these sources, we initially identified 11 GCMs as EC-Earth3, EC-Earth3-Veg-LR, CNRM-CM6-1, MIROC6, MPI-ESM1-2-HR, GFDL-CM4, GFDL-ESM4, KIOST-ESM, INM-CM4-8, ACCESS-CM2, and HadGEM3-GC31-LL. To minimize uncertainty associated with regional model mismatch, we refined our selection to five GCMs that demonstrated the best performance for the study area. Model performance was evaluated by comparing historical precipitation, maximum, and minimum temperature data from each GCMs against corresponding observed data using average values (1970 to 2014).

Model performance was evaluated using three statistical measures i.e., percentage bias (Pbias), Pearson’s correlation coefficient, and root mean square error (RMSE). These metrics were computed for monthly precipitation, maximum, and minimum temperature. A rank-based evaluation was then applied to determine the five best-performing models based on their overall statistical performance. The BCSD bias-correction method employed in NEX-GDDP-CMIP6 has been shown to effectively reduce systematic biases in both temperature and precipitation, particularly improving the representation of seasonal patterns and extreme values relevant for drought assessment 33 , 35 , 36 , 37 . Our model evaluation using monthly data (1970–2014) is appropriate for validating GCM performance at the temporal scales relevant to our SPEI-based drought analysis.

Although GCM selection was limited to the 1980–2014 period due to the availability of historical climate model runs up to 2014, all subsequent climate change and drought analyses utilized the complete observational record (1980–2024) to establish the most robust baseline for future projections and to capture recent drought conditions. To assess future changes in precipitation, temperature, and drought conditions across the study area, we defined a temporal framework that contains the historical baseline (1980–2024) with two future projection periods: near future (2026–2055) and far future (2056–2085). These periods were selected following standard practice in climate impact assessments and IPCC-based studies. The baseline period represents the recently observed climate under current warming conditions and provides a robust reference for anomaly and trend analysis. The near-future period (2026–2055) corresponds to the mid-21st century, which is highly relevant for short- to medium-term adaptation planning, infrastructure design, and agricultural management. The far-future period (2056–2085) represents the late-21st century. In addition, we considered 30 years period for this division to reduce natural variability.

We employed the SPEI as the primary drought indicator for this assessment. Originally developed by Vicente-Serrano et al. 21 , SPEI was designed to incorporate temperature-driven hydrological processes in drought quantification by accounting for both atmospheric water supply (precipitation) and atmospheric water demand (potential evapotranspiration, PET). The index was formulated as an enhancement to the Standardized Precipitation Index (SPI) developed by McKee et al. 40 .

SPEI offers several methodological advantages over the traditional SPI. The inclusion of PET allows SPEI to capture the influence of rising temperatures that lead to increased evapotranspiration rates and subsequently exacerbated drought conditions 41 . SPEI calculation is based on the monthly water balance (D), defined for month i as:

where P represents precipitation and PET represents potential evapotranspiration. For this study, SPEI calculations were performed using the Climpact software package, which utilizes the SPEI R package implementation. SPEI calculation in R package is estimated using the Thornthwaite (1948) method. We input daily precipitation, maximum and minimum temperature and latitude and longitude of the stations. For Harris County-wide analyses, spatially averaged values were computed for all meteorological variables, with representative latitude and longitude coordinates calculated as the centroid of all grid points within the county boundaries.

Drought events were identified using the run theory approach 42 with a fixed SPEI threshold. The value of SPEI ≤ −0.8 was used as a threshold for drought initiation and termination, following 43 , 44 . This threshold corresponds to moderate drought conditions and enables comprehensive drought detection suitable for assessing regional impacts.

Run theory identifies continuous sequences of months with SPEI values below the specified threshold, with each sequence representing a distinct drought event. In this study, for each event, two drought characteristics were calculated: (1) drought duration, defined as the number of consecutive months with SPEI < − 0.8, and (2) drought severity, defined as the cumulative deficit represented by the sum of absolute SPEI values over the drought period. Thereafter, we developed the SDF curves.

SDF curves were constructed by categorizing drought events according to their duration and examining the corresponding severity values within each category. When a duration category contained adequate drought events, the severity distributions were fitted to three probability distributions frequently employed in hydrological frequency analysis: Gamma, Weibull, and Log-normal. The best-fitting distribution for each duration category was chosen based on the Akaike Information Criterion (AIC), with lower AIC values indicating superior model fit while penalizing model complexity 45 . This approach achieves an optimum balance between goodness-of-fit and model simplicity. For each drought duration and corresponding best-fitting distribution, drought severities corresponding to different return periods were assessed using:

where T is the return period (2, 5, 10, 20, 50 years) and p is the exceedance probability. Quantiles of the fitted distribution were extracted using the inverse cumulative distribution function to provide severity values at each return period. The resulting SDF dataset represents the expected drought severity for different durations and return periods under each climate scenario. We adopted the basic concepts of drought duration, severity, intensity, and SDF curves given by Cavus and Aksoy 46 .

The Mann-Kendall (MK) test was employed to detect temporal trends in drought indices and meteorological variables throughout the study period. This non-parametric statistical test is widely applied in meteorological and hydrological time series analysis 29 , 37 due to its robustness and versatility.

The Mann-Kendall test evaluates the null hypothesis of no trend against the alternative hypothesis of either an increasing or decreasing monotonic trend. However, the test indicates only the direction of trends, not their magnitude. Therefore, Sen’s slope estimator was used in conjunction with the MK test to quantify the magnitude of detected trends. The MK test statistic S is calculated as:

Where the sign function is defined as

where X j and X i values represent sequential data values, and n is the number of observations in the time series. A positive S value indicates an ‘upward trend’, and a negative value of S indicates a ‘downward trend’.

For n > 10, the test statistic S , is approximately normally distributed with a mean equal to zero and the variance of S is computed as:

For n > 10 the standard normal variate z is computed using the following equation:

Thus, in a two sided test for trend, the null hypothesis is rejected if \(\left| z \right|> z_{{\alpha /2}}\) .

Hydroclimatic time series often exhibit serial correlation, which can lead to erroneous trend detection by inflating the probability of identifying significant trends where none exist 47 , 48 . To address this limitation, we implemented the modified Mann-Kendall test proposed by Hamed and Rao 48 .

This modified approach accounts for autocorrelation by adjusting the variance calculation using an effective sample size, thereby providing more reliable trend detection in the presence of serial correlation. The method maintains robustness while correcting for the artificial inflation of significance that can occur with correlated data.

Statistical significance was evaluated using a two-sided test at the α = 0.05 significance level. Under this criterion, standardized test statistics |z| > 1.96 indicate statistically significant trends, with positive values representing increasing trends and negative values representing decreasing trends

Following the methodology outlined and guided by relevant literature for the study region, an initial pool of 11 GCMs was used for evaluation. The model evaluation was conducted using spatially averaged values across Harris County rather than individual grid-point comparisons. This approach was adopted because spatial aggregation better captures the regional precipitation variability patterns and provides more robust statistical relationships with observed data series. Grid-point-based evaluations revealed low correlation values for precipitation (> 0.4), despite accurately capturing temperature variations with the observed data. Table 1 presents the evaluation results, including individual statistical parameters, model-specific ranks for each metric, and the corresponding rank sum values. Based on this evaluation framework, five GCMs were identified as best suited for drought assessment in Harris County. They are EC-Earth3, EC-Earth3-Veg-LR, CNRM-CM6-1, MIROC6, and ACCESS-CM2.

These selected models represent diverse modeling approaches and institutional frameworks, ensuring a robust ensemble that captures model structural uncertainty while maintaining high performance standards for the study region. Averaging the daily values from each of these 5 GCMs, we formulated a mean ensemble, which was mainly used for drought assessment.

To understand the overall changes in climate conditions, we assessed the percentage changes in precipitation and absolute changes in temperature relative to the baseline period (1980–2024). These changes were calculated for the ensemble average of the selected GCMs across three SSP scenarios, and the results are presented in Table 2 . Across all scenarios and time periods, future projections indicate decreased precipitation and increased temperatures throughout Harris County.

On an annual scale, the most substantial precipitation decline is projected under SSP2-4.5 for the 2055–2085 period, with an average decrease of 16.8%. Temperature increases are most pronounced under the SSP5-8.5 scenario, with the highest maximum and minimum temperature changes recorded as 4.13 °C and 3.95 °C, respectively, for the 2055–2085 time slice under SSP5-8.5.

Considering the monthly scale, temperature increases are consistently higher under the SSP5-8.5 scenario across all months, with the most pronounced changes projected for the far future period. Monthly precipitation changes exhibit greater variability across different months compared to temperature. The most severe precipitation reduction (−35.95%) is projected for August under the SSP5-8.5 scenario during the 2055–2085 period. Moreover, the most severe temperature increment, both maximum and minimum, was recorded in November (Table 2 ).

Figures 2 and 3 illustrate the temporal evolution of future annual precipitation, maximum, and minimum temperature across the five selected GCMs and their ensemble average over Harris County for different scenarios spanning 2026–2085. The results demonstrate consistent patterns across all models: decreasing precipitation trends and increasing temperature trends (both maximum and minimum) throughout the projection period. However, notable inter-model variability is observed in the magnitude of these climate parameter changes, highlighting the importance of using multiple GCMs in the ensemble approach.

Projected precipitation variation over Harris County under ensemble mean and individual GCMs for different SSP scenarios.

Projected maximum ( a , b , c ) and Minimum ( d , e , f ) temperature variation over Harris County under ensemble mean and individual GCMs for different SSP scenarios.

Projected SPEI-12 drought index for Harris County under three emission scenarios: ( a ) SSP1-2.6, ( b ) SSP2-4.5, and ( c ) SSP5-8.5 for the period 2026-2085

Temporal variation of historical SPEI-6 and SPEI-12 averaged over Harris County

The SPEI-12 time series averaged over Harris County revealed distinct drought patterns under different climate scenarios (Fig. 4 ). We observed approximately the same variations of SPEI-6, but with different magnitudes. Drought events are identified using the established threshold of SPEI ≤ −0.8, as indicated in the methodology. Figure 5 presents the historical baseline from 1980 to 2024 of SPEI-6 and SPEI-12, providing essential context for future projections.

Both historical observations and future projections across all SSP scenarios (SSP1-2.6, SSP2-4.5, and SSP5-8.5) demonstrate recurrent drought incidents of varying intensity and duration for both SPEI-6 and SPEI-12 indices. The SPEI values exhibit a typical range of + 2.5 to −2.5, capturing the full spectrum from extreme wet conditions (potential flooding) to severe drought events.

Scenario-specific drought patterns demonstrate contrasting trajectories across emission pathways. Under SSP1-2.6 (Fig. 4 a), drought occurrence remains relatively stable throughout the projection period, with drought events distributed relatively evenly between the near-term (2026–2055) and far-future (2056–2085) periods. The moderate emission pathway SSP2-4.5 (Fig. 4 b) shows a gradual increase in drought frequency and intensity, particularly after mid-century. The high-emission scenario SSP5-8.5 (Fig. 4 c) exhibits the most pronounced intensification, with drought events becoming both more frequent and more severe in the latter half of the 21 st century. SPEI-6 analysis (not shown but analyzed) reveals similar temporal patterns with different magnitudes, reflecting the shorter accumulation period’s sensitivity to short-term precipitation variability.

A notable trend emerges from the future SPEI projections: drought events show a marked intensification toward the latter half of the 21 st century across most of the GCMs under all scenarios (Fig. 4 ). To quantify this temporal evolution, drought frequencies were calculated as percentages for four distinct periods: historical early (1980–2002) and recent (2003–2024), and future near-term (2026–2055) and long-term (2056–2085) periods (Table 3 ). The analysis reveals a pronounced increase in drought frequency during the far-future period (2056–2085) across ensemble mean, EC-Earth3, and ACCESS-CM2 under all scenarios. This trend is particularly pronounced under the high-emission SSP5-8.5 scenario, where all GCMs project drought frequencies exceeding twice those of the near-future period (2026–2055) (Table 3 ).

Interestingly, the historical analysis indicates a temporary reduction in drought frequency during the recent period (2003–2024) compared to the earlier historical period (1980–2002). This creates a distinctive temporal pattern characterized by initially high drought occurrence (1980–2002), followed by a relative decrease (2003–2024), and subsequently a substantial increase in future projections, particularly in the far-future period. This oscillating pattern suggests that Harris County may be experiencing a transitional period of relatively lower drought frequency, which is projected to reverse dramatically under continued greenhouse gas emissions. The magnitude of this projected increase varies with emission scenarios, with SSP5-8.5 showing the most severe drought frequencies.

Trend analysis was conducted on SPEI-6 and SPEI-12 drought indicators across all climate scenarios using five GCMs and their ensemble means. Due to the presence of lag-one autocorrelations in the time series, the modified Mann-Kendall test was primarily employed to ensure robust statistical inference. Trend analysis was performed for the complete future period (2026–2085) and historical period (1980–2024), with results summarized in Table 4 . The trend analysis reveals distinct patterns depending on emission pathways and their associated warming levels:

Considering the medium and high emission scenarios (SSP2-4.5 and SSP5-8.5), both SPEI-6 and SPEI-12 exhibit statistically significant negative trends ( p < 0.05) across the ensemble mean and most individual GCMs, with the notable exception of CNRM-CM6-1. These negative trends indicate a consistent shift toward drier conditions and increased drought severity throughout the 21 st century under moderate to high greenhouse gas emission pathways. Under the Paris Agreement-consistent low emission pathway (SSP1-2.6), the trend patterns are markedly different 49 . The ensemble means, EC-Earth3-Veg-LR, ACCESS-CM2, and MIROC6 show non-significant positive trends for both SPEI-6 and SPEI-12, characterized by very small Sen’s slopes (approximately 0.0001). However, the remaining GCMs still project negative trends, suggesting model-dependent responses to limited warming scenarios.

The historical trend analysis (1980–2024) provides important context for interpreting future projections. Both SPEI-6 and SPEI-12 demonstrate slight positive trends during the historical period, indicating a modest wetting tendency. Notably, the SPEI-12 trend reaches statistical significance ( p < 0.05) with a Sen’s slope of 0.0008, suggesting a slight increase in moisture availability over longer timescales during recent decades.

The apposition of historical and future trends reveals a fundamental shift in the hydroclimate trajectory of Harris County. The transition from positive historical trends (wetting tendency) to predominantly negative future trends (drying tendency) represent a critical inflexion point in regional water resources. This transition is most pronounced under higher emission scenarios, where the negative trends are both statistically robust and consistently projected across multiple climate models. The magnitude of this hydroclimatic shift emphasizes the sensitivity of regional drought patterns to greenhouse gas emission pathways, with SSP1-2.6 offering the potential to maintain near-stable conditions while higher emission scenarios project substantial drying trends.

These contrasting trends highlight the importance of emission pathway selection for regional drought risk management. The statistical significance and consistency of drying trends under SSP2-4.5 and SSP5-8.5 scenarios suggest high confidence in projected drought intensification, while the SSP1-2.6 scenario offers a pathway toward climate stabilization with minimal drought trend intensification.

A threshold of − 0.8 SPEI (applied to both 6-month and 12-month time scales) was used to define the onset and termination of drought events. For each identified drought event, drought severity and drought duration were determined, and drought intensity was further calculated.

Drought severity is expressed as the cumulative sum of the absolute SPEI values over the drought period, while drought intensity was defined as severity divided by duration, representing the average magnitude of drought conditions per month. Figure 6 illustrates the variation of drought severity under SPEI-6 and SPEI-12, while Fig. 7 presents the variation of drought intensity. Both figures indicate that drought severity and intensity generally increase with longer drought durations.

From the SPEI-6 severity scatter plot, it is evident that longer-duration droughts tend to have higher severities under historical conditions, whereas shorter-duration droughts exhibit greater severity under future climate projections. In contrast, this tendency is less distinct in the SPEI-12 scatter plot, where historical and future events appear more mixed. However, SPEI-12 clearly shows that high-emission scenarios, SSP2-4.5 and SSP5-8.5, produce higher drought severities compared to lower-emission scenarios (SSP1-2.6). The SPEI-12 drought intensity scatter plot also reflects this pattern: SSP1-2.6 demonstrates the lowest intensity, while SSP2-4.5 and SSP5-8.5 exhibit higher values. Moreover, the variability of drought intensity is greater in future scenarios than in the historical baseline, highlighting the increased uncertainty and potential severity of future droughts.

The study formulated SDF curves for return periods of 2, 5, 10, 20, and 50 years. Drought durations were determined based on the identified drought events in each dataset, with different durations emerging for various scenarios. The results are presented for ensemble GCM projections and compared with historical data. Both SPEI-6 and SPEI-12 indices were analyzed, as shown in Fig. 8 .

The SDF curves demonstrate the relationship where higher return periods correspond to higher drought severity, confirming that severe drought events occur less frequently than moderate ones. SPEI-6 captured a wider range of drought durations compared to SPEI-12, reflecting its sensitivity to medium-term drought conditions versus the longer-term focus of SPEI-12. Across all SDF curves (all scenarios, SPEI-6 and SPEI-12), drought severity consistently increased with increasing duration, which is an expected and fundamental characteristic of drought behavior 46 , 50 . This relationship reflects the cumulative nature of drought impacts, where longer drought events accumulate greater deficits over time, resulting in higher total severity values.

Scenario comparisons revealed contrasting patterns between the two indices. Considering SPEI-6, the historical drought severity exceeded future scenario projections, particularly for drought durations longer than 2 months. For shorter durations (1–2 months), severity values were comparable between historical and future scenarios. In contrast, for SPEI-12, future scenarios showed significantly higher drought severity than historical conditions, with severity values approximately 5 times greater than those observed in the historical period, indicating substantial intensification of long-term drought conditions under climate change.

We observed comparatively higher drought severity under SPEI-6 in historical time periods (1980 to 2024) due to the number of severe droughts that actually happened during the period, including droughts in 2009, 2011 to 2014 51 , 52 .

It should be noted that SDF curves were constructed using probability distribution fitting for drought durations with sufficient events (≥ 3) to ensure statistical reliability. This approach captures recurrent drought patterns effectively but may exclude rare extreme drought events that occur infrequently in the projection period. Such extreme events, while statistically unreliable for frequency analysis, may represent critical tail risks that warrant separate consideration in adaptation planning.

Variation of drought severity with duration under different scenarios for ensemble mean and historical data using SPEI-6 and SPEI-12.

Variation of drought intensity with duration under different scenarios for ensemble mean and historical data using SPEI-6 and SPEI-12.

Drought Severity-Duration-Frequency (SDF) curves for return periods of 2, 5, 10, 20, and 50 years under different scenarios for ensemble mean and historical data over Harris County.

The drought assessment was performed using NEX-GDDP-CMIP6, a statistically downscaled and bias-corrected dataset widely recognized for its reliability in future climate and extreme event assessments 36 , 53 . Consistent with the methodological framework proposed by Li et al. 16 , we excluded climate models characterized as “hot models” due to their equilibrium climate sensitivity exceeding 5 °C. We used only the best-fit 5 GCMs with normal climate sensitivity in order to obtain realistic, plausible future scenarios, and we captured the structural uncertainty of diverse models using these five GCMs: EC-Earth3, EC-Earth3-Veg-LR, CNRM-CM6-1, MIROC6, and ACCESS-CM2.

This study revealed that the combined interactions of future reductions in precipitation and continued increases in temperature will result in more frequent droughts in Harris County. We also noted that future precipitation changes show greater variability among models, while temperature increases are consistent across different models, scenarios, and time slices. Our analysis also demonstrated that there are substantial variations in precipitation reductions over different months, which warrants further research into the shifting of rainy seasons. Under all scenarios and time slices, the August precipitation reduction is substantial (−35.95% under the SSP5-8.5 scenario during the 2055–2085 period) compared to the historical reference period (1980 to 2024). Our findings of reduced August precipitation in Harris County align with broader regional studies reporting general precipitation declines and summertime reductions across Texas 14 , 54 .

This study demonstrated higher precipitation variability through both graphical representation (Fig. 2 ) and SPEI-6 and SPEI-12 index analysis (Fig. 4 ). The higher variability in these indices implies the occurrence of not only droughts but also floods, as evidenced by elevated positive SPEI values in some years. Future flood severity may increase due to projected increases in precipitation intensity, which are attributed to a warmer Gulf of Mexico supplying more moisture to the lower atmosphere 55 , 56 . However, while recognizing this compound risk, this study focuses primarily on drought assessment. Both indices (SPEI-6 and SPEI-12) indicate projections of more frequent drought occurrences compared to the reference period. Comparable to the present findings, Li et al. 16 , also showed a net water availability reduction of about 6% during 2045–2054 and a 4% reduction during 2065–2074 under high emission scenarios in the Houston area of Texas using CMIP6 data.

The present analysis indicates the likely occurrence of more frequent droughts in Harris County during the far future (2056–2085), as evidenced by our drought frequency and trend analyses (Tables 3 and 4 ). Particularly, the study identified a distinctive three-phase pattern: high drought frequency (1980–2002), temporary reduction (2003–2024), and projected intensification (2056–2085). Trend analysis over the complete study periods corroborates these findings: the historical period (1980–2024) exhibits a slight positive SPEI trend despite early high drought frequency, reflecting the influence of the recent reduction phase (2003–2024), while future projections (2026–2085) show significant negative trends under medium and high emission scenarios, with the far-future intensification (2056–2085) driving this overall drying trajectory. This pattern suggests that Harris County may currently be experiencing a transitional period of relatively lower drought risk. However, the projected doubling of drought frequency under SSP5-8.5 represents a critical threshold for regional water security, requiring heightened attention from policymakers for proactive planning and adaptation strategies.

The projected doubling of drought frequency under high-emission scenarios and statistically significant decreasing trends in SPEI have significant implications for regional water management and drought preparedness strategies. The temporal clustering of drought events toward the end of the century suggests that adaptation measures will become increasingly critical for maintaining water security in Harris County. These findings align with broader regional climate projections for the southern United States, where increasing temperatures and altered precipitation patterns are expected to decrease water availability, particularly under high-emission pathways 16 .

Furthermore, with continuous urbanization in Harris County, the urban heat island effect may amplify temperature-driven drought intensification. This study identifies this as a research gap for future investigation on how the urban heat island affects future drought conditions in Harris County.

This study employed SPEI-6 and SPEI-12 to capture short-term and long-term drought dynamics, respectively. Despite showing consistent future drought trends, these indices revealed scale-dependent drought characteristics: SPEI-6 indicated greater severity during the historical period compared to future projections, while SPEI-12 displayed approximately the opposite behavior, with future drought severity substantially exceeding historical conditions. This contrasting pattern has important implications for different water management sectors, with short-term operational planning facing relatively stable conditions while long-term water security planning must prepare for substantial intensification.

SDF curves provide a quantitative basis for infrastructure design and drought preparedness planning. The 5-fold increase in SPEI-12 severity for future scenarios necessitates re-evaluation of existing infrastructure design criteria and agricultural planning. Research findings of this study are crucial for updating reservoir storage capacities, water supply networks, and drought contingency plans. Under high emission scenarios, this 5-fold increase in SPEI-12 drought severity has significant implications for reservoir sizing and demand management in the county. The results also show a considerable increase in drought severity under SPEI-6 for durations exceeding two months compared to historical conditions, which can inform phased adaptation strategies.

The projected increase in drought frequency and severity reported in the study results from the combined and interactive effects of reduced precipitation and enhanced evaporative demand under warming conditions. These temperature increases directly enhance atmospheric evaporative demand through increased potential evapotranspiration, which is a fundamental component of SPEI calculation. Even with modest precipitation changes, elevated temperatures can shift the precipitation-evapotranspiration balance toward drought conditions 21 .

The pronounced August precipitation reduction (−35.95% under SSP5-8.5, 2055–2085) coincides with the warmest months, creating a compounding effect where reduced moisture supply occurs simultaneously with peak evaporative demand. This timing is particularly critical for Harris County, as August represents a critical period for water resource stress when irrigation demands peak and reservoir levels are typically at seasonal lows. Our findings align with broader regional studies indicating that future drought intensification in the southern United States is driven by temperature-induced increases in evaporative demand rather than precipitation decreases alone 14 .

The three-phase temporal pattern we identified likely reflects the interaction between anthropogenic forcing and natural climate variability. The recent reduction in drought frequency (2003–2024) may be influenced by decadal-scale modes of climate variability such as the Atlantic Multidecadal Oscillation (AMO), Pacific Decadal Oscillation (PDO), El Nino-Southern Oscillation index which modulate regional precipitation patterns 57 – 59 . However, the projected future intensification demonstrates that anthropogenic warming dominates over natural variability under continued greenhouse gas emissions.

Moreover, Harris County’s location along the Gulf Coast makes it particularly sensitive to changes in Gulf of Mexico sea surface temperatures, which influence both moisture availability for precipitation and atmospheric evaporative demand. Future studies should employ coupled climate-land surface models to quantify how urbanization and land use change interact with climate change to modify drought characteristics in Harris County through urban heat island effects and altered surface water balances.

Policymakers can utilize the transitional period (2026–2055), characterized by lower drought frequency, as an opportunity for building resilience before far-future intensification. This window allows for testing and refining adaptation strategies, investing in infrastructure upgrades, and developing institutional capacity.

The current study has several important limitations. The study used NASA NEX-GDDP-CMIP6 data, which uses the Bias-Correction Spatial Disaggregation (BCSD) method 33 , 34 as a bias-correction method. While BCSD has been widely applied, the method has several well-documented limitations. Specifically, BCSD assumes stationarity of model biases 60 , may underestimate changes in variability and extremes 61 , and does not explicitly account for physical land–atmosphere feedback processes 62 . Moreover, spatial drought analysis was not conducted because point-based GCM historical data showed poor agreement with observed reference period data. Moreover, NEX-GDDP-CMIP6’s spatial resolution provided only 10 grid points over Harris County. Despite this coarse resolution, the results are applicable county-wide due to Harris County’s relatively small size (4,600 km²) and flat topography. Another limitation involves the SDF analysis threshold requiring ≥ 3 drought occurrences, which ensures statistical reliability but may omit rare extreme events that could pose significant risks.

This study assessed future drought conditions in Harris County, Texas, using five CMIP6 GCMs under three SSP scenarios (SSP1-2.6, SSP2-4.5, and SSP5-8.5) for 2026–2085. The analysis employed SPEI-6 and SPEI-12 indices to characterize drought dynamics at multiple temporal scales, providing comprehensive insights into future drought risk in this rapidly urbanizing region.

Projections indicate substantial increases in drought frequency, severity, and intensity, particularly during 2056–2085. Combined precipitation reductions (up to 35.95% in August under SSP5-8.5) and temperature increases drive a projected doubling of drought frequency under high-emission pathways, representing a critical threshold for regional water security.

Analysis revealed a distinctive pattern: high drought frequency (1980–2002), temporary reduction (2003–2024), and projected intensification (2056–2085). This suggests Harris County currently experiences a transitional period of relatively lower drought risk that will reverse sharply by mid-century. Trend analysis corroborates this pattern, with historical positive SPEI trends (wetting tendency) transitioning to significant negative trends under SSP2-4.5 and SSP5-8.5 scenarios.

Under SSP1-2.6, most models project non-significant trends, suggesting potential climate stabilization. Conversely, SSP2-4.5 and SSP5-8.5 scenarios show statistically significant negative trends ( p < 0.05), emphasizing the critical role of emission mitigation in managing drought risk.

Severity-Duration-Frequency analysis reveals that future SPEI-12 droughts will be five times more severe than historical events across all return periods (2–50 years). This necessitates comprehensive re-evaluation of reservoir storage capacities, water supply networks, and agricultural planning frameworks, as current infrastructure designed for historical conditions may prove inadequate under future climate scenarios.

These findings provide critical information for proactive drought mitigation strategies in Harris County and serve as a reference for drought risk assessment across the southern United States.

All relevant data included in the paper are available in repositories of NEX-GDDP-CMIP6 and Daymet.

Abeysingha, N. & Rajapaksha, U. SPI-based spatiotemporal drought over Sri Lanka. Adv. Meteorol. 2020 (1), 9753279 (2020).

Ahmad, L. et al. Drought concepts, characterization, and indicators. Integrated drought management, volume 1. CRC Press. 43–62. (2023).

Tikuye, B. G., Ray, R. L., Manjunatha, B., Tefera, G. W. & Gurau, S. 2025 Drought monitoring using the Climate Hazards InfraRed Precipitation with Stations (CHIRPS) in Ethiopia. Nat. Hazards Res. 5 , 348–362 (2025).

Rahman, G., Jung, M.-K., Kim, T.-W. & Kwon, H.-H. Drought impact, vulnerability, risk assessment, management and mitigation under climate change: A comprehensive review. KSCE J. Civ. Eng. 29 (1), 100120 (2025).

Sun, J. et al. Will drought evolution accelerate under future climate?. J. Hydrol. 650 , 132552 (2025).

Han, J. & Singh, V. P. A review of widely used drought indices and the challenges of drought assessment under climate change. Environ. Monit. Assess. 195 (12), 1438 (2023).

Moon, S. & Ha, K.-J. Future changes in monsoon duration and precipitation using CMIP6. npj Clim. Atmos. Sci. 3 (1), 45 (2020).

Veettil, A. V., Fares, A. & Awal, R. Winter storm Uri and temporary drought relief in the western climate divisions of Texas. Sci. Total Environ. 835 , 155336 (2022).

Wang, S. S., Zhao, L., Yoon, J.-H., Klotzbach, P. & Gillies, R. R. Quantitative attribution of climate effects on Hurricane Harvey’s extreme rainfall in Texas. Environ. Res. Lett. 13 (5), 054014 (2018b).

Li, Z., Li, X., Wang, Y. & Quiring, S. M. Impact of climate change on precipitation patterns in Houston, Texas, USA. Anthropocene 25 , 100193 (2019).

Ray, R. L., Fares, A. & Risch, E. Effects of drought on crop production and cropping areas in Texas. Agric. Environ. Lett. 3 (1), 170037 (2018).

Seager, R. et al. Projections of declining surface-water availability for the Southwestern United States. Nat. Clim. Chang. 3 (5), 482–486 (2013).

Cook, B. I., Ault, T. R. & Smerdon, J. E. Unprecedented 21st-century drought risk in the American Southwest and Central Plains. Sci. Adv. 1 (1), e1400082 (2015).

Article ADS PubMed PubMed Central Google Scholar

Nielsen-Gammon, J., Escobedo, J., Ott, C., Dedrick, J. & Van Fleet, A. Assessment of historic and future trends of extreme weather in Texas, 1900–2036. (2020).

Tebaldi, C. et al. Climate model projections from the Scenario Model Intercomparison Project (scenariomip) of CMIP6. Earth Syst. Dyn. 12 (1), 253–293 (2021).

Li, W., Li, D., El-Askary, H., Fisher, J. B. & Fang, Z. N. Evaluating future water availability in Texas through the lens of a data-driven approach leveraged with CMIP6 general circulation models. Sci. Total Environ. 921 , 171136 (2024).

Liu, W., Feddema, J., Hu, L., Zung, A. & Brunsell, N. Seasonal and diurnal characteristics of land surface temperature and major explanatory factors in Harris County. Tex. Sustain. 9 (12), 2324 (2017).

Hoque, N., Granato, J. & McNeil, C. Projections of the population of Texas and counties in Texas by age, sex, and race/ethnicity from 2010 to 2050. Houston, Texas. Retrieved from http://www.uh.edu/class/hobby/_docs/research/population/2014 PPRLE-SV2 (2014).

Tirivarombo, S. & Eliasson, O. P. Drought monitoring and analysis: Standardised Precipitation Evapotranspiration Index (SPEI) and Standardised Precipitation Index (SPI). Phys. Chem. Earth Parts A B C 106 , 1–10 (2018).

Seneviratne, S. I. et al. Weather and climate extreme events in a changing climate. (2021).

Vicente-Serrano, S. M., Beguería, S. & López-Moreno, J. I. A multiscalar drought index sensitive to global warming: The standardized precipitation evapotranspiration index. J. Clim. 23 (7), 1696–1718 (2010).

Tian, L., Yuan, S. & Quiring, S. M. Evaluation of six indices for monitoring agricultural drought in the South-central United States. Agric. For. Meteorol. 249 , 107–119 (2018).

Henson, A. & Rosenthal, J. The importance of investing in policies beyond ‘the basics’: Perspectives from black fathers in Philadelphia on how to increase community safety. Crit. Soc. Policy . 43 (3), 536–556 (2023).

Awal, R., Fares, A. & Habibi, H. Optimum turf grass irrigation requirements and corresponding water-energy-CO2 nexus across Harris County, Texas. Sustainability 11 (5), 1440 (2019).

Statkewicz M. D., Talbot R, Rappenglueck B. Changes in precipitation patterns in Houston, Texas. Environ. Adv. 5 , 100073 (2021)

Chun B, Hur M, Won J. Impacts of Thermal Environments on Health Risk: A Case Study of Harris County, Texas. Int. J. Environ Res Public Health. 2021 21;18(11):5531. https://doi.org/10.3390/ijerph18115531 .

Ramage, J. K. & Adams, A. C. Scientific Investigations Report 2025–5028, Status of Water-Level Altitudes and Long-Term and Short-Term Water-Level Changes in the Chicot and Evangeline (Undifferentiated) and Jasper Aquifers, Greater Houston Area, Texas, https://doi.org/10.3133/sir20255028 2024.

Thornton, P. E. et al. Gridded daily weather data for North America with comprehensive uncertainty quantification. Sci. Data . 8 (1), 190 (2021).

Article PubMed PubMed Central Google Scholar

Abeysingha, N., Arachchige, S. M., Ray, R. L. & Muthuwatta, L. Variability of rainfall and the cropping season of a tropical dry zone river basin: A case study in the Malwathu Oya River Basin in Sri Lanka. Adv. Meteorol. 2025 (1), 2967930 (2025).

Feng D, Liu J, Lawson K, Shen C. Differentiable, Learnable, Regionalized Process-Based Models With Multiphysical Outputs can Approach State-Of-The-Art Hydrologic Prediction Accuracy. https://doi.org/10.1029/2022WR032404 (2022).

Johnson Z C, Snyder C D, Hitt N P. Landform features and seasonal precipitation predict shallow groundwater influence on temperature in headwater streams. https://doi.org/10.1002/2017WR020455 (2017)

Tefera GW, Dile YT, Srinivasan R, Baker T, Ray RL (2023) Hydrological modeling and scenario analysis for water supply and water demand assessment of Addis Ababa city, Ethiopia. J Hydrol Reg Stud 46:101341

Thrasher, B. et al. NASA global daily downscaled projections, CMIP6. Sci. Data . 9 (1), 262 (2022).

Wood, A. W., Leung, L. R., Sridhar, V. & Lettenmaier, D. P. Hydrologic implications of dynamical and statistical approaches to downscaling climate model outputs. Clim. Change 62 (1), 189–216 (2004).

Y, B. et al. Future changes in extremes across China based on the Nex-GDDP-CMIP6 models. Clim. Dyn. 62 (10), 9587–9617 (2024).

Ghazi, B., Salehi, H., Przybylak, R. & Pospieszyńska, A. Assessment of drought conditions under climate change scenarios in Central Europe (Poland) using the standardized precipitation index (SPI). Clim. Serv. 39 , 100591 (2025).

Jiang, F., Wen, S., Gao, M. & Zhu, A. Assessment of nex-gddp-cmip6 downscale data in simulating extreme precipitation over the Huai River basin. Atmosphere 14 (10), 1497 (2023).

Yang, Y. L. Navigating climate models; understanding future climate Trends in south central Texas, Native Prairies Association of Texas, https://texasprairie.org/navigating-climate-models-understanding-future-climate-trends-in-south-central-texas/ 2024.

Wootten, A. et al. Customized statistically downscaled CMIP5 and CMIP6 projections: Application in the Edwards Aquifer Region in South-Central Texas. Earth’s Future 12 (10), e2024EF004716 (2024).

McKee, T. B., Doesken, N. J. & Kleist, J. The relationship of drought frequency and duration to time scales. Proc. 8th Conf. Appl. Climatol . 17 , pp179–183 (1993). California.

Nwayor, I. J. & Robeson, S. M. Exploring the relationship between SPI and SPEI in a warming world. Theor. Appl. Climatol. 155 (4), 2559–2569 (2024).

Yevjevich, V. M. An objective approach to definitions and investigations of continental hydrologic droughts (Colorado State University, 1967).

Das, J., Jha, S. & Goyal, M. K. Non-stationary and copula-based approach to assess the drought characteristics encompassing climate indices over the Himalayan states in India. J. Hydrol. 580 , 124356 (2020).

Merz, N. & Zachariah, M. Understanding droughts under climate change in South America based on severity-duration-frequency curves and drought atlases. Clim. Change . 178 (9), 163 (2025).

Bozdogan, H. Model selection and Akaike’s information criterion (AIC): The general theory and its analytical extensions. Psychometrika 52 (3), 345–370 (1987).

Cavus, Y. & Aksoy, H. Critical drought severity/intensity-duration-frequency curves based on precipitation deficit. J. Hydrol. 584 , 124312 (2020).

Abeysingha, N., Singh, M., Sehgal, V., Khanna, M. & Pathak, H. Analysis of rainfall and temperature trends in the Gomti River Basin. J. Agric. Phys. 14 (1), 56–66 (2014).

Hamed, K. H. & Rao, A. R. A modified Mann-Kendall trend test for autocorrelated data. J. Hydrol. 204 (1–4), 182–196 (1998).

Wang, F. et al. Global and regional climate responses to national-committed emission reductions under the Paris Agreement. Geogr. Ann. A Phys. Geogr. 100 (3), 240–253 (2018a).

Aksoy, H. et al. Critical drought intensity-duration-frequency curves based on total probability theorem-coupled frequency analysis. Hydrol. Sci. J. 66 (8), 1337–1358 (2021).

McGregor, K. M. Comparison of the recent drought in Texas to the drought of record using reanalysis modeling. Papers in Applied Geography 1 (1), 34–42 (2015).

Nielsen-Gammon, J. & McRoberts, B. An assessment of the meteorological severity of the 2008-09 Texas drought through July 2009. (2009).

Gummadi, S., Samineni, S. & Lopez-Lavalle, L. A. B. Assessing high-resolution precipitation extremes in Central Asia: Evaluation and future projections. Clim. Change 178 (2), 29 (2025).

Bukovsky, M. S., McCrary, R. R., Seth, A. & Mearns, L. O. A mechanistically credible, poleward shift in warm-season precipitation projected for the US Southern Great Plains?. J. Clim. 30 (20), 8275–8298 (2017).

Easterling, D. R. et al. Precipitation change in the United States. (2017).

Maloney, M. C. & Preston, B. L. A geospatial dataset for US hurricane storm surge and sea-level rise vulnerability: Development and case study applications. Clim. Risk Manag. 2 , 26–41 (2014).

Barlow, M. et al. A 2019 North American extreme precipitation events and related large-scale meteorological patterns: a review of statistical methods, dynamics, modeling, and trends. Clim. Dyn. 53 , 6835–6875 (2019).

Nowbuth, M. D. Assessing climate variability using extreme rainfall and temperature indices. Univ. Maurit. Res. J. 16 (1), 196–206 (2010).

Nielsen-Gammon, J. W., 2011: The 2011 Texas Drought: A Briefing Packet for the Texas Legislature. https://hdl.handle.net/1969.1/158245

Dixon, K. W. et al. Evaluating the stationarity assumption in statistically downscaled climate projections: Is past performance an indicator of future results?. Clim. Change 135 , 395–408. https://doi.org/10.1007/s10584-016-1598-0 (2016).

Gutmann, E. et al. An intercomparisonof statistical downscaling methodsused for water resource assessments inthe United States. Water Resour. Res. 50 , 7167–7186. https://doi.org/10.1002/2014WR015559 (2014).

Berg, A., Findell, K., Lintner, B. et al. Land–atmosphere feedbacks amplify aridity increase over land under global warming. Nature Clim Change 6, 869–874 (2016). https://doi.org/10.1038/nclimate3029 .

This work was supported by the Natural Resources and Conservation Services (NRCR), United States Department of Agriculture (USDA) [grant number: NR237442XXXXC023].

College of Agriculture, Food and Natural Resources, Prairie View A&M University, Prairie View, TX, 77446, USA

N. S. Abeysingha, Ram L. Ray, Kusalika M. Kularathna & D. M. B. M. Deegala

Department of Agricultural Engineering and Soil Science, Faculty of Agriculture, Rajarata University of Sri Lanka, Anuradhapura, Sri Lanka

Search author on: PubMed Google Scholar

N.S. Abeysingha developed the methodology, conducted formal analysis, and drafted the original manuscript. Ram L. Ray contributed to methodological development and supported in reviewing and editing the manuscript. D.M.B.M. Deegala and Kusalika M Kularathna assisted with data collection and tabulation. All authors read and approved the final manuscript.

The authors declare no competing interests.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ .

Abeysingha, N.S., Ray, R.L., Kularathna, K.M. et al. Drought pattern under climate change in Harris County, Texas, USA based on CMIP6 projections. Sci Rep 16 , 14948 (2026). https://doi.org/10.1038/s41598-026-42547-3

DOI : https://doi.org/10.1038/s41598-026-42547-3

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

&copy; 2026 Springer Nature Limited

Join the Conversation