Annotated Bibliography for GEOG 560 GIScience I: Introduction to Geographic Information Science

Suryabhagavan, K.V., 2017. GIS-based climate variability and drought characterization in Ethiopia over three decades. Weather and climate extremes, 15, pp.11-23.

Geographic Information Systems (GIS) is a very useful tool for agriculture and natural resource management. Spatial-interpolation techniques such as inverse distance weighted (IDW), Spline, and Kriging interpolation that are available in ArcGIS can be used for data reconstruction. In this study, the author aims to understand the climate variability over three decades (1983–2012) in the potential crop-growing regions of Ethiopia. Daily rainfall and temperature (maximum and minimum) data were obtained for 120 weather stations of the National Meteorological Agency (NMA) and Ethiopian Institute for Agricultural Research (EIAR), and 87 out of the 120 stations were selected based on the data length and completeness. Mann–Kendall test was used on the annual and seasonal distribution of rainfall and temperature to detect the presence of trends in the climate data, the data’s temporal variability was calculated using mean, standard deviation, and coefficient of variation (CV%). Then, Standardized Precipitation Index (SPI) was used to study the drought frequency and severity over Ethiopia with only rainfall as the input data. Regional drought was studied using the spatial pattern of at-site (point) droughts. Various drought periods were identified across Ethiopia during the study period, including major droughts that occurred during 1983–1984 and 2002–2003 and affected approximately two-thirds of the study area. Findings from the study can be useful for improving the regional drought monitoring in Ethiopia and neighboring countries. Additionally, the use of remote sensed data and GIS techniques can contribute a key role in drought monitoring, prediction, and preparedness.

Satti, S.R. and Jacobs, J.M., 2004. A GIS-based model to estimate the regionally distributed drought water demand. Agricultural Water Management, 66(1), pp.1-13.

Water demand in the future in states like Florida, that require permits for consumptive uses of water, involves the inclusion of new consumptive uses as well as modifications of existing permitted consumptive uses as a response to land-use changes. To support regional planning in such contexts, the authors have developed a GIS-based model called GIS-based Water Resources and Agricultural Permitting and Planning System (GWRAPPS). GWRAPPS was developed within the ArcGIS framework by coupling the Agricultural Field Scale Irrigation Requirements Simulation (AFSIRS) crop water model with ArcGIS. GWRAPPS can be used to simulate the irrigation requirements of thousands of farms growing 60 varieties of crops grown on the 766 soil types in Florida using 9 different irrigation systems. The authors demonstrated the use of the model in estimating crop water requirements in Volusia County, FL using 2 different approaches: firstly, using the predominant soil type in the farm and later using all the soils in the farms. The model results were dependent on the farm location, soil, and climate parameters. The study found the soil texture variability to be important at the local scale (farm) but not significant on a regional scale. From the data requirement point of view, GWRAPPS uses minimal data from permit applicants. The model acts as a decision support system (DSS) for permitting and planning irrigation water demand in future drought scenarios.

KHAMPEERA, A., YONGCHALERMCHAI, C. and TECHATO, K., 2018. Drought monitoring using drought indices and GIS techniques in Kuan Kreng peat swamp, Southern Thailand. Walailak Journal of Science and Technology (WJST), 15(5), pp.357-370.

Drought can be felt in various forms and is studied by using different drought indices. Kuan Kreng Peat Swamp (KKPS) in southern Thailand suffers from peat fires and water shortages resulting from the degradation of the peatland. This study aims to use a Geographical Information System (GIS) Spatial Analysis to identify the areas of KKPS susceptible to drought. The authors have used Standardized Precipitation Index (SPI) to assess meteorological drought, Normalized Difference Drought Index (NDDI) to assess vegetative drought, and Standardized Water Level Index (SWI) and Water Table Level (WTL) to assess hydrological drought. They focused on two years: 2010 (moderate El Nino year) and 2012 (non-El Nino year) and studied forest fires throughout the study period. A flowchart of the steps used in this study is provided below:

The authors found observed severe drought in all 5 zones of the peat swamps in 2010 and less severe drought in 2012, with hydrological drought occurring in almost all areas in July, August, and September. In the El Niño year 2010, the monthly SPI maps of the study area showed that the most severe drought occurred during the dry season and the area was under drought stress throughout the year. They also found that meteorological drought (SPI) and hydrological drought (WTL and SWI) are strongly correlated with rainfall amount and drought could occur even during normal rainfall due to factors such as water use for irrigation and draining of the area. Findings from this study can be used to study drought hazards in KKPS, assess the risk of peat fires, and identify fire-vulnerable areas.

Asbury, Z., 2018. A Geospatial Study of the Drought Impact on Surface Water Reservoirs: Study Cases from Texas and California. University of Arkansas.

Texas and California are two states very familiar with the impacts of drought. This study uses remote sensing and geographic information system (GIS) to study the impact of drought on surface water reservoirs in San Angelo (5 lakes) and Dallas (5 lakes) in Texas, and Lake Oroville in California. Changes in reservoir sizes during the summer and winter months during the years 2005 – 2016 were quantified and correlated against local climate using Landsat-5, -7, and -8 multispectral imageries obtained from the Earth Resources Observation and Science (EROS) data center. GIS-based density slicing was employed to slice the raster Near Infrared (NIR) images for all reservoirs into three categories: deep water, shallow water, and dry area. The accuracy of the classified images was checked and overall accuracy of >93% was found. The study found that Lake Oroville was affected mostly by the changes in precipitation, and drought played a major role in the shrinkage and expansion of the surface water reservoirs. Extreme drought in San Angelo caused several small lakes to disappear and were no longer used for fishing or recreation.

Bahrami, M., Bazrkar, S. and Zarei, A.R., 2021. Spatiotemporal investigation of drought pattern in Iran via statistical analysis and GIS technique. Theoretical and Applied Climatology, 143(3), pp.1113-1128.

Reconnaissance Drought Index (RDI) is based on precipitation and evapotranspiration, can help identify different drought types, and is sensitive to a variety of climate conditions. This study aims to evaluate the spatiotemporal characteristics of drought in Iran on annual and seasonal scales. It uses 48 years (1967–2014) of data from 40 synoptic meteorological stations of Iran covering all climatic conditions. FAO Penman-Monteith (FAO-56) equation was used to estimate the Potential evapotranspiration (PET). Shapiro-Wilk’s statistical test was used to assess the normality of Standardized RDI, and the IDW interpolation technique in ArcGIS 10.2 was used to prepare the spatial distribution maps for drought. The results indicated an increasing trend in drought, with “most upward trend” observed in winter (for seasonal) and annual time scales and “least upward trend” in summer. The study also found that coverage of extreme, severe, and moderate drought showed a positive trend, and the highest portion of the study area was under moderate drought.

Chopra, P., 2006, January. Drought risk assessment using remote sensing and GIS: a case study of Gujarat. Enschede, The Netherlands: ITC.

Drought is a common phenomenon in the Indian State of Gujarat, with the state experiencing 43 years of drought between 1901 – 2000. In this study, the author focuses on the use of Remote Sensing and Geographic Information Science in Drought risk evaluation by identifying areas facing agricultural and meteorological drought using the Normalized Difference Vegetation Index (NDVI) based on NOAA-AVHRR (8km) images and Standardized Precipitation Index (SPI). The monthly rainfall of 169 rain stations was used to analyze the relation between NDVI and rainfall and to derive SPI using 23 years of data (1981 – 2003). The methodology used in calculating the drought risk in the study is given below:

Inverse Distance Weighted (IDW) was used to interpolate the SPI values using ArcGIS 9.0 and the interpolated raster layer was reclassified into 6 drought severity classes: no drought, abnormally dry, moderate drought, severe drought, extreme drought, and exceptional drought. The study found a positive correlation between rainfall and NDVI, including a strong correlation in water-limiting areas. The study also found that central and northeastern parts of Gujarat were more prone to drought than other areas. Similarly, SPI showed significant relation with NDVI anomaly and food-grain yield anomaly, suggesting that SPI can be used as regional crop production and vegetation status indicator.

Hammouri, N. and El-Naqa, A., 2007. Drought assessment using GIS and remote sensing in Amman-Zarqa basin, Jordan. Jordan J Civ Eng, 1(2), pp.142-152.

Amman-Zarqa basin in Jordan is important from both an agricultural and hydrological perspective. The rainy season in the basin, which starts every year in October and ends in April, has seen quite a variability. The authors aimed to assess whether the basin was facing drought conditions or not. In doing so, they used rainfall data and satellite images to compute the Standardized Precipitation Index (SPI) and Normalized Difference Vegetation Index (NDVI) and evaluate drought. The study used GIS software to create a spatial digital database of meteorological records, generate thematic layers of drought for SPI and NDVI, and delineate areas with high drought risks. From the SPI results, they found that the dry period in the basin lasts between 3 – 5 years followed by a wet period of 2 – 3 years. From NDVI analysis, they asserted that drought was dominant in some months such as December and October, for wet conditions were dominant for November and January.

Prathumchai, K., Honda, K. and Nualchawee, K., 2001, November. Drought risk evaluation using remote sensing and GIS: a case study in Lop Buri Province. In 22nd Asian conference on remote sensing (Vol. 59).

Thailand has been experiencing an increase in drought frequency in recent years. Droughts are often underestimated as their onset is slow and has less visual impact. This study evaluates the drought risk using remote sensing and GIS in Lop Buri Province, Thailand. It identifies drought risk areas, based on modified Ministry of Science, Technology and Environment (MOSTE) criteria by studying the decrease in Normalize Difference Vegetation Index (NDVI) in a drought year. Based on the annual rainfall of Lop Buri province for 1988-1998, two years were identified for further analysis: 1994 (drought year) and 1996 (normal year), and satellite images were obtained. Then, the difference between NDVI computed for two years was calculated. Modified MOSTE drought risk area was computed based on 7 factors: annual average rainfall, rainy days frequency, annual evaporation, irrigation area, groundwater resources, slope, and soil drainage. The predominant drought level in Lop Buri based on modified MOTSE criteria was identified to be “high drought risk” comprising approximately 51% of the area, mostly concentrated in North, North-West, and South-West parts of the study area. It was observed that over two-thirds of the land area saw a decrease in NDVI in the drought year (1994) compared to the normal year (1996). It was found that forest areas had the highest decrease in NDVI, whereas crops and paddy fields had a lower decrease in NDVI. There was a negatively correlated relationship between drought risk and NDVI change that was significant only for the first three drought risk levels (very high, high, and moderate).

Abdulrazzaq, Z.T., Hasan, R.H. and Aziz, N.A., 2019. Integrated TRMM data and standardized precipitation index to monitor the meteorological drought. Civil Engineering Journal, 5(7), pp.1590-1598.

Drought is a major problem in the semi-arid and arid regions of Iraq, however, drought research and monitoring require an extensive record of climate data. In poor and data-scarce regions, remotely sensed satellite-based rainfall such as the Tropical Rainfall Measuring Mission (TRMM) can be of valuable importance in understanding the drought. In this study, the authors used TRMM precipitation data to analyze the meteorological drought at 11 stations in Western Iraq based on the Standardized Precipitation Index (SPI) for the years 2000, 2005, 2010, 2015, and 2017. The study area is characterized by lack of rainfall, absence of vegetation and is considered a large rocky desert plateau. The steps used in the drought analysis by the authors is given below:

SPI values were interpolated using the Kriging method to identify the drought areas. Results indicate that droughts were higher than expected, particularly in the South and South-West regions of the study area. The authors proposed four categories (mild, moderate, severe, and extreme) of drought classification in Iraq. They conclude that TRMM data can be used in absence of observed data due to their high accuracy, wide-coverage, and easily accessible nature.

Palchaudhuri, M. and Biswas, S., 2016. Application of AHP with GIS in drought risk assessment for Puruliya district, India. Natural Hazards, 84(3), pp.1905-1920.

Multi-criteria assessment using analytical hierarchy process (AHP) is a popular method of using geographic information system (GIS) in disaster risk management, particularly for floods and landslide hazards. In this study, the authors extend the use of AHP to spatial analysis of drought in West Bengal, India. They used 14 different parameters: annual rainfall, monthly rainfall, maximum temperature, monthly temperature, maximum evapotranspiration, monthly evapotranspiration, relative humidity, soil texture, land use/land cover, slope, groundwater, cultivators, agricultural laborers, and population as factors influencing the drought and created a thematic map for each factor using GIS. Then, AHP was used to calculate the weightage of each of the 14 factors, based on which drought severity map was created. ArcView Model Builder was used to execute the weighted overlay for the integration of input data layers to create the drought-related output layer. The study found that 70 % of the area of Puruliya district in West Bengal is under severe drought. The drought severity map helps understand the drought vulnerability of the study area and guides local stakeholders to make proper decisions to mitigate the impacts of the drought.

Rahman, M.R. and Lateh, H., 2016. Meteorological drought in Bangladesh: assessing, analysing and hazard mapping using SPI, GIS and monthly rainfall data. Environmental Earth Sciences, 75(12), pp.1-20.

The study analyzes meteorological drought characteristics of Bangladesh using the standardized precipitation index (SPI) and geographic information system (GIS). It uses monthly rainfall for the period 1971–2010 for 34 meteorological stations distributed all over Bangladesh and obtained from the Bangladesh Meteorological Department (BMD) to compute SPI at 3- and 6-month time scales. Inverse distance weighted (IDW) interpolation in GIS was used to study the spatial extent of drought. An integrated drought hazard index (DHI) was computed using the analytical hierarchy process (AHP) with drought at different time scales and severity as inputs to generate a drought hazard map. Finally, 4 levels of drought hazards were defined based on the index values as low (zone I), moderate (zone II), high (zone III), and very high (zone IV). The study reported that there were 10 most drought-affected years (1972, 1978, 1981, 1982, 1995, 1997, 1999, 2004, 2006, and 2010) and 6 worst drought years based on severity (1972, 1978, 1982, 1995, 1999, and 2006). They also found that the frequency of moderate drought was higher than severe and extreme drought. The areas identified as vulnerable for drought include northern, north-western, western, south-western, and central parts of Bangladesh and were affected by low annual and seasonal rainfall, high rainfall variability, climate change, and increase in maximum temperature. The authors recommended “urgent intervention on a priority basis” to mitigate the impacts of drought in these areas as it impacts the agriculture and water sector the most.

Zagade, N.D. and Umrikar, B.N., 2021. Drought severity modeling of upper Bhima river basin, western India, using GIS–AHP tools for effective mitigation and resource management. Natural Hazards, 105(2), pp.1165-1188.

Upper Bhima River Basin in Western India has experienced severe and prolonged droughts in the recent past and is a major concern for stakeholders in the region. This study develops a multi-parameter-based analytical hierarchy process (AHP) model to understand the spatial extent of drought in the study area and includes 4 sub-basins of the upper Bhima river basin. 10 parameters affecting the drought vulnerability of the Bhima river basin have been used in the study: NDVI, rainfall, slope, vadose zone, soil depth, LULC, water harvesting structures, geomorphology, drainage density, and groundwater level fluctuation. The relative weightage of each parameter to the drought vulnerability was calculated using the AHP methodology. Results from the study show that 24% area of the study area has severe drought and 31% of the area has moderate drought vulnerability. These areas are largely agriculture-based with high water demand and experience drinking water scarcity. This drought vulnerability information can be useful for drought preparedness and mitigation measures in the affected villages of the study area.

Bezdan, A., Benka, P., Grabic, J. and Salvai, A., 2012. Estimation of agricultural drought vulnerability using GIS tools: A case study of Vojvodina Region (Serbia). Proceedings of the Balwois.

Serbian agriculture is dependent on the timing of the precipitation. Vojvodina region, lying in northern Serbia, is a predominantly agricultural region where drought can have a negative impact on the production of crops. The case study analyzes the agricultural drought vulnerability of the study area using a GIS-based method. The authors identified 5 key factors affecting agricultural drought in the region: climate, soil properties, land use, geomorphology, and access to irrigation and assigned a numerical weight ranging from 1 to 5 to the identified factors. They computed a 3-month SPI using the precipitation data from 1971 to 2010. Drought vulnerability maps were generated by summing up the 5 thematic layers and reclassifying the resulting layer into six classes of agricultural drought vulnerability. The study estimated that 13.2% area in the Vojvodina region is highly vulnerable to agricultural drought, located mostly in the central and southeastern parts of the region. The authors conclude that thematic and drought vulnerability maps generated in this study could be used to inform decision-makers in the study area of the hazard and implement mitigation techniques to reduce the impacts of drought on agriculture.

Wu, D., Qu, J.J. and Hao, X., 2015. Agricultural drought monitoring using MODIS-based drought indices over the USA Corn Belt. International Journal of Remote Sensing, 36(21), pp.5403-5425.

The corn belt in the United States, comprising of midwestern states were hit by a severe drought in 2012. This study focuses on the analysis of Moderate Resolution Imaging Spectroradiometer (MODIS) time series for the years 2000–2012 to assess the 2012 drought during the corn-growing season. The study area included the US states of Iowa, Illinois, Indiana, Michigan, Ohio, Nebraska, Minnesota, and Missouri. The MODIS data were resampled to surface reflectance data with a resolution of 500 m. Cropland Data Layer obtained from USDA NASS was used in the analysis as well. Daily precipitation data from 129 weather stations were used to compute the SPI time series and 13 years (2000–2012) of monthly Palmer-Z data were obtained for the same stations from the North American Drought Monitor (NADM) database. USDM data was used to identify the Drought events and classify normal and drought years in the study area. Seven MODIS indices (NDVI, NDWI, NDII6, NMDI, VCI, VHI, and PDI) for crops pixels were averaged over normal years to calculate the normal growing conditions for corn, and current crop conditions were assessed by comparing MODIS indices with normal conditions. The flowchart of the methodology used in this study is given below:

MODIS-based drought index anomalies were compared with SPI and Palmer-Z and MODIS drought maps were compared with USDM maps. Results from the study indicate that MODIS index anomalies correlated better with six-month SPI, suggesting the agricultural drought conditions due to rainfall deficiency. MODIS index NDII6 showed correlated best with SPI-6 and Palmer-Z, indicating good sensitivity to medium-term precipitation stress and moisture deficiency in dense crop areas. The authors suggest NDII6 has a strong potential for application as an agricultural drought monitoring remote-sensing index.

Schwarz, M., Landmann, T., Cornish, N., Wetzel, K.F., Siebert, S. and Franke, J., 2020. A spatially transferable drought hazard and drought risk modeling approach based on remote sensing data. Remote Sensing, 12(2), p.237.

In this study, the authors have developed a drought hazard model that has a higher spatial resolution than most available drought models and is spatially transferable to other regions. First, a logistic regression model to predict drought was developed for rangelands and

croplands in the USA, and results were verified with that of the United States Drought Monitor (USDM). Then, the accurate USA model was transferred and calibrated for South Africa and Zimbabwe where drought hazard, drought vulnerability, and drought risks were assessed. The data used in this study include land use, crop yield, precipitation, surface reflectance, LST, and albedo. Additionally, population density, gross domestic product, farming systems, and livestock density information were also used for drought vulnerability and risks study. The steps used in the study are as given below:

The Drought Hazard model was first implemented in Missouri Basin, and then successfully transferred to South Africa and Zimbabwe after calibrating and setting it up. The model has a good predictive quality for the USA and South Africa, but only moderate predictive quality for Zimbabwe. Overall, with the help of qualitative analyses, the authors found a good match between the model results and the Global Drought Observatory results. The authors claim the applicability of the model over different regions by changing input variables, weights, and crops affected by drought depending on the salient characteristics of the area.

Additional Readings:

1. Abatzoglou, J.T., McEvoy, D.J. and Redmond, K.T., 2017. The west wide drought tracker: drought monitoring at fine spatial scales. Bulletin of the American Meteorological Society, 98(9), pp.1815-1820.
2. Belal, A.A., El-Ramady, H.R., Mohamed, E.S. and Saleh, A.M., 2014. Drought risk assessment using remote sensing and GIS techniques. Arabian Journal of Geosciences, 7(1), pp.35-53.
3. Fallahati, A., Soleimani, H., Alimohammadi, M., Dehghanifard, E., Askari, M., Eslami, F. and Karami, L., 2020. Impacts of drought phenomenon on the chemical quality of groundwater resources in the central part of Iran—Application of GIS technique. Environmental monitoring and assessment, 192(1), pp.1-19.

Contact: Sudip Gautam