Dalia Grendaitė

Ms. Dalia Grendaitė

Vilnius University, Institute of Geosciences
orcid

Research Focus & Metrics

Academic Career Length

9Years (2018-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
28 March 2022 in Remote Sensing
Verified
Journal Article
Open access
PDF file
ESEdvinas Stoneviĉius
Edvinas Stoneviĉius
GUGiedrius Uselis
Giedrius Uselis
DGDalia Grendaitė
Dalia Grendaitė
Climate change leads to more variable meteorological conditions. In many Northern Hemisphere temperate regions, cold seasons have become more variable and unpredictable, necessitating frequent river ice observations over long sections of rivers. Satellite SAR (Synthetic Aperture Radar)-based river ice detection models have been successfully applied and tested, but different hydrological, morphological and climatological conditions can affect their skill. In this study, we developed and tested Sentinel-1 SAR-based ice detection models in 525 km sections of the Nemunas and Neris Rivers. We analyzed three binary classification models based on VV, VH backscatter and logistic regression. The model sensitivity and specificity were used to determine the optimal threshold between ice and water classes. We used in situ observations and Sentinel-2 Sen2Cor ice mask to validate models in different ice conditions. In most cases, SAR-based ice detection models outperformed Sen2Cor classification because Sen2Cor misclassified pixels as ice in areas with translucent clouds, undetected by the scene classification algorithm, and misclassified pixels as water in cloud or river valley shadow. SAR models were less accurate in river sections where river flow and ice formation conditions were affected by large valley-dammed reservoirs. Sen2Cor and SAR models accurately detected border and consolidated ice but were less accurate in moving ice conditions. The skill of models depended on how dense the moving ice was. With a lowered classification threshold and increased model sensitivity, SAR models detected sparse frazil ice. In most cases, the VV polarization-based model was more accurate than the VH polarization-based model. The results of logistic and VV models were highly correlated, and the use of VV was more constructive due to its simpler algorithm.

11 July 2018 in Geologija. Geografija
Verified
Journal Article
Open access
PDF file
DGDalia Grendaitė
Dalia Grendaitė
ESEdvinas Stonevičius
Edvinas Stonevičius
JKJūratė Karosienė
Jūratė Karosienė
KSKsenija Savadova
Ksenija Savadova
JKJūratė Kasperovičienė
Jūratė Kasperovičienė
Inland waters are an important habitat for flora and fauna and are also used for aesthetic, recreational, and industrial needs; therefore, monitoring the current state of freshwaters and applying measures to improve water quality are of high importance. To have an efficient monitoring system that could cover large areas, the use of remote sensing data is crucial. In this study the suitability of the Sentinel-2 Multispectral Imager data is tested for observing cyanobacteria bloom events in the eutrophic lakes and retrieving the chlorophyll-a concentration – an indicator of phytoplankton biomass. The analysis is carried out using data from four lakes in Lithuania – two eutrophic blooming lakes and two oligo-mesotrophic non-blooming lakes. The results showed that reflectances are higher in the eutrophic lakes than in the oligo-mesotrophic lakes due to the presence of an optically active constituent, namely, chlorophyll-a pigment. We tested empirical equations for chlorophyll-a concentration retrieval in eutrophic lakes derived in other studies to check whether they could be used without adaptation to local conditions. Most of the equations performed well (R2 = 0.5–0.8); however, they had high RMSEs = 17–53 μg L–1. The equation used with the bottom of atmosphere data CHL8_L2A (R2 = 0.76) had the lowest RMSE = 9 μg L–1. In addition, we derived empirical equations for eutrophic lakes in Lithuania. The equations that were based on the Sentinel-2 band ratio B5/B4 and the three band (B4, B5, and B8A) expression performed the best (R2 = 0.77–0.79) and had lower RMSE = 7 μg L–1 than empirical equations from other studies. A larger in situ dataset could improve the algorithm performance in retrieving the chlorophyll-a concentration. The first attempts to map water quality parameters in eutrophic lakes in Lithuania using the data received from the Sentinel-2 MSI sensor show good results, as the changes in reflectance, caused by the changes in chlorophyll-a concentration, can be seen from satellite images.

28 July 2021 in Geocarto International
Verified
Journal Article
Closed access
DGDalia Grendaitė
Dalia Grendaitė
ESEdvinas Stoneviĉius
Edvinas Stoneviĉius
One of the largest uncertainties in remote sensing data comes from atmospheric influence. This research aims to explain the uncertainties emanating from atmospheric correction (AC) product selection and how they influence chlorophyll α concentration retrieval in lakes in eastern Lithuania. We tested seven products from six AC processors (Acolite, Acolite Rayleigh, iCOR, Sen2Cor, C2RCC, C2X, and POLYMER) and 10 chlorophyll α retrieval algorithms with different architectures. The uncertainty of AC products transferred to chlorophyll α concentrations, and large differences in the chlorophyll α concentrations retrieved using different AC products were observed. The match-up analysis showed that chlorophyll α algorithms based on band difference performed best in terms of a high coefficient of determination and the lowest median bias when used with image-based, Sen2Cor, and TOA data. The results of this study highlight the uncertainties of AC products as well as how the selection of the chlorophyll α retrieval algorithm can mitigate the influence of AC selection.

12 January 2025 in Water Research
Verified
Journal Article
Closed access
Ksenija Savadova-RatkusKsenija Savadova-Ratkus
Ksenija Savadova-Ratkus
DGDalia Grendaitė
Dalia Grendaitė
JKJūratė Karosienė
Jūratė Karosienė
ESEdvinas Stoneviĉius
Edvinas Stoneviĉius
JKJūratė Kasperovičienė
Jūratė Kasperovičienė
JKJudita Koreivienė
Judita Koreivienė

14 February 2024 in IEEE Access
Verified
Journal Article
Open access
DGDalia Grendaitė
Dalia Grendaitė
Linas PetkevičiusLinas Petkevičius
Linas Petkevičius
Algal blooms are a common problem in inland waters, which raise growing awareness on monitoring lakes’ conditions. The on site monitoring is expensive and requires large human resources efforts. This work proposes remote monitoring techniques using satellite images and machine learning algorithms to predict chlorophyll concentration in water bodies and identify algal blooms. The training and test dataset used in this study includes diverse range of lakes in Baltic countries. The lake spectral features obtained from Sentinel-2 satellite images are used as predictors for proposed deep neural network models. The prediction of chlorophyll concentration with MAE 7.97 mg/m 3 and bloom vs. non-bloom classification with 71.6 % accuracy was achieved. The use of Bèzier curves for smoothing the point-wise prediction is proposed for identification of algal bloom characteristics: the bloom start date, end date, and duration. The results showed lake type impact on the blooming time. The experimental data and code are released with paper.

28 May 2022 in Water
Verified
Journal Article
Open access
PDF file
DGDalia Grendaitė
Dalia Grendaitė
ESEdvinas Stoneviĉius
Edvinas Stoneviĉius
Inland waters are dynamic systems that are under pressure from anthropogenic activities, thus constant observation of these waters is essential. Remote sensing provides a great opportunity to have frequent observations of inland waters. The aim of this study was to create a data-driven model that uses a machine learning algorithm and Sentinel-2 data to classify lake observations into four biophysical classes: Clear, Moderate, Chla-dominated, and Turbid. We used biophysical variables such as water transparency, chlorophyll concentration, and suspended matter to define these classes. We tested six machine learning algorithms that use spectral features of lakes as input and chose random forest classifiers, which yielded the most accurate results. We applied our two-step model on 19,292 lake spectra for the years 2015–2020, from 226 lakes. The prevalent class in 67% of lakes was Clear, while 19% of lakes were likely affected by strong algal blooms (Chla-dominated class). The models created in this study can be applied to lakes in other regions where similar lake classes are found. Biophysical lake classification using Sentinel-2 MSI data can help to observe long-term and short-term changes in lakes, thus it can be a useful tool for water management experts and for the public.