Jessica Besnier
PhD Candidate-Hydrologist-Environmental Engineer​
​
Research
Current Research Interests:
Hydrology
Hydrological Extremes
Machine Learning
Remote Sensing
Water Resource
Allocation
Data Visualization
Research Projects:
In 2019, there was an inception of a drought in the La Plata River Basin (LPRB) that lasted until
2021. With the LPRB being the second largest river basin in South America, this drought
affected all aspects of life for over 100 million people who inhabit this area. The most
detrimental effects were to the water resources, hydropower, and agriculture sectors. In this
study, we utilize satellite datasets to examine spatial and temporal variability of hydrological
anomalies throughout the basin. These variable anomalies, such as precipitation and groundwater, are derived from the Global Land Data Assimilation System (GLDAS) and were found to decrease from 2019 to 2021. The water loss was calculated from 2015 to 2021 using the Mann-Kendall test to assess the changes in anomalies for the terrestrial water storage throughout the basin and subbasins. It was found that the Upper Parana subbasin lost the most water over this period. Additionally, the Soil Moisture Active Passive (SMAP) observations were used in the spatial and temporal analysis and characterization of the drought. These datasets were compared and validated with in-situ observations. These products allowed for an improved understanding of the spatial variability within the entire basin and sub basins of the LPRB. This analysis facilitates understanding of droughts in this basin as well as resources to assist in the future management of water resources.
Using GRACE TWS to predict reservoir height in the Upper Parana River Basin
All over the world, water levels are constantly changing. From lakes to rivers to oceans, the patterns of the water levels change due to different factors. With hydrological extremes increasing in intensity and duration around the world, it is important to understand what changes these levels in order to better predict and mitigate the negative impacts of changing water levels. We use estimates of terrestrial water storage (TWS) variability from the Gravity Recovery and Climate Experiment (GRACE) satellite missions to predict reservoir operation in Brazil. To do this, reservoir water elevations are derived from multi-satellite radar altimetry (RA) data and used as a proxy of their operation. 16 reservoirs in Southern Brazil are analyzed.
For each reservoir, the Pettitt test was used to identify the point break within the TWS data, and the Mann-Kendall test was used to identify trends before and after these breaks. A machine learning approach was used to reconstruct RA-based water elevations using GRACE data. The approach considered numerous geomorphologic and meteorologic characteristics of reservoir including precipitation (from GPM IMERG) and temperature (GLDAS). For some of the reservoirs, various ML models were run with a 5-day forecast horizon and the outputs were compared to determine which model predicted most accurately for each reservoir. Some of the models incorporated in this study include decision tree regressor, kernel ridge model, linear regression, random forest regression, and support vector regression. The findings of this study will give insight into what variables affect the relationship be-tween TWS and RA height in the Upper Parana Basin in Southern Brazil to improve prediction measures for reservoir height.
Understanding the Inniscarra reservoir fluctuations to predict downstream flooding through hydrological modeling
Cork, Ireland is prone to flooding; the city has experienced more than 300 major floods over the past two centuries. To mitigate floods in the city, the Inniscarra dam was built between 1953-1957 about 14km upstream from the city. In the past decade, there have been 3 separate flood events costing 90 million, 40 million, and 30 million Euros in damage. To improve flood protection and reduce damage costs, it is necessary to understand the previous flooding scenarios. Water level data in the reservoir upstream has implications for Cork and surrounding areas.
The goal of this project is to develop Soil &Water Assessment Tool (SWAT) and Machin Learning (ML) techniques to understand and predict the downstream river discharge. This will provide insight into how the fluctuations in the upstream reservoir are predictive of downstream river discharge, and thus flooding in Cork City. Algorithms for this project are created by using climate variables as inputs and the algorithm predicts the fluctuations in terms of the discharge forecast as output. Multiple ML algorithms, such as random forest regression, support vector machine, and decision tree regressor, will be created to compare and determine the best predictive output. The discharge outputs from the ML models will be used to compare the results for the historical period with the SWAT model outputs. Overall, this project works to understand the fluctuations of the reservoir in order to predict and mitigate future flooding.
Earth observations give insight into the correlation between drought and food security
Droughts cause over 25% of the world’s population to face food insecurity. Droughts decrease water resources, damage ecosystems, reduce hydropower production, limit food production, and hinder vital economic industries including agriculture, transportation, and recreation. There are countless inequalities in food security, as the effects of climate change disproportionately affect developing countries and lower income areas. These regions do not have adequate tools to predict and mitigate damage from these natural disasters. With this, climate change is predicted to increase droughts in coming years, thus indicating even more need for drought observations.
From 2019-2021, there was a severe drought in the La Plata River Basin (LPRB) in South America, drastically affecting the agriculture, hydropower, and water resource sectors. Based on previous research, the subbasins most adversely affected by the drought are the upper Parana, upper Uruguay, and lower Uruguay basins. Historically, when droughts occur, there is little to no precipitation which reduces agricultural yield, leading to food shortage and food insecurity.
Currently, one of the major shortcomings in drought prediction models is predicting the location, magnitude, and assistance needed when the disaster strikes. This project determines the drought variability, impacts of drought on food security on the catchment level, and changes in food shortage due to climate change in the Upper Parana and Uruguay river basins to focus on addressing the gaps in current drought prediction models. By looking at the relationship between different aspects of the water cycle, the combination of Earth observation satellite products will allow for an assessment of drought variability over the span of the 2019-2021 drought. Various Machine Learning Algorithms will be used to assist in predicting the location, magnitude, and assistance needed in droughts.
In this work we investigate the land surface response to the 2015-2016 drought in the Lower Mekong River Basin for five different watersheds using a combination of in-situ and observations from NASA satellite sensors. The analysis uses the observed streamflow, precipitation from the Global Precipitation Mission (GPM), soil moisture (9km) from the Soil Moisture Active Passive (SMAP) sensor, downscaled 1km SMAP soil moisture, and terrestrial water storage from the Gravity Recovery and Climate Experiment (GRACE). The results on the temporal propagation of drought and the recovery shows interesting patterns. Specifically, the use of the 1km soil moisture has advantages in understanding the spatial heterogeneity, specifically in smaller watersheds where the precipitation (at 10km) is at coarse spatial resolution. In performing the analysis, we use lagged correlations between soil moisture and streamflow, precipitation and streamflow, precipitation and total water storage, as well as the indices of Standardized Precipitation Index (SPI) and Standardized Streamflow Index (SSI).