Predictive mapping of aquatic ecosystems by means of support vector machines and random forests

Martínez‐Santos, Pedro; Aristizábal, H.F.; Díaz-Alcaide, S.; Gómez-Escalonilla, Víctor

doi:10.1016/j.jhydrol.2021.126026

Cited by 17 publications

(7 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Second, adopting a similar approach to the one proposed by Zimmerman et al (2018), we performed a method that combines the 10‐fold cross‐validation and the RF–recursive feature elimination analysis (RFE, see Pedregosa et al, 2011), so that we can balance the construction of the best‐performing model and the risk of overfitting. RF‐RFE is aimed at selecting features by recursively considering smaller and smaller sets of features (Martínez‐Santos et al, 2021). Specifically, RF algorithm is first trained on the entire set of input variables and the importance of each of these is assessed by the impurity‐based criterion.…”

Section: Methodsmentioning

confidence: 99%

Shedding light on the decline of Iberian freshwater fish species over the period 1980–2020

et al. 2022

View full text Add to dashboard Cite

1. Freshwater fish biodiversity is experiencing an alarming decline worldwide.Understanding the main factors behind its deterioration is a key step for ecosystem restoration. In this work, large-scale and long-term data were used to identify the causes of the decline of native species richness in Castilla-La Mancha.This region in central Spain covers part of six river basins belonging to four of the 11 biogeographical provinces for freshwater fish in the Iberian Peninsula.2. Firstly, we built a dataset that associates the presence of several fish species and a wide range of environmental variables (e.g. hydrological and hydromorphological indicators, land use classes, presence of alien fish species) at selected river sites for two different time periods (1980-2000 and 2001-2020). Secondly, we conducted an exploratory data analysis to identify possible temporal trends in the dataset. Finally, we applied the random forest algorithm to predict the response of different ecological guild-based metrics of fish richness to the selected variables.3. The exploratory data analysis revealed a decrease in native fish species richness in 74% of the area studied. There was no sustained temporal trend for stressor variables, except for the number of alien species, which increased in most river sites (63%). The models of the richness of native rheophilic, native intolerant, alien rheophilic, and alien limnophilic species performed satisfactorily. Magnitude of maximum discharge, presence of alien species, land use in the catchment area and altitude were the most important predictors of richness of native intolerant and rheophilic species. Alien limnophilic species proved to be sensitive to variables related to flow regime alteration, such as the presence of dams and the number of river flow reversals, while a less degraded habitat was found to be favourable to alien rheophilic species.4. The results suggest that the cumulative effect of persistent altered flow regimes and water pollution, coupled with a strong increase in the number of alien species, have led to the decline of native species in the area studied. The restoration This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.

show abstract

Section: Methodsmentioning

confidence: 99%

Shedding light on the decline of Iberian freshwater fish species over the period 1980–2020

et al. 2022

View full text Add to dashboard Cite

show abstract

“…This can affect the performance of the classifiers by attributing extra weight to an input variable or by adding noise to the final outcomes. Interpretability can also be impaired because the regression coefficients of certain algorithms are not uniquely determined (Martínez-Santos et al, 2021).…”

Section: Supervised Classification Routinementioning

confidence: 99%

“…While useful, these are of limited value in cases where the input dataset consists solely of unambiguous examples. Furthermore, there are question marks as to whether these metrics are truly representative for the development of spatially-distributed estimates (Martínez-Santos et al, 2021). In those instances, using ad hoc calibration elements, such as complementary field information, can contribute to a better interpretation of the outcomes.…”

Section: Introductionmentioning

confidence: 99%

Preprocessing approaches in machine learning-based groundwater potential mapping: an application to the Koulikoro and Bamako regions, Mali

Gómez-Escalonilla

Martínez‐Santos

Martín-Loeches

2021

Preprint

View full text Add to dashboard Cite

Abstract. Groundwater is crucial for domestic supplies in the Sahel, where the strategic importance of aquifers can only be expected to increase in the coming years due to climate change. Groundwater potential mapping is gaining recognition as a valuable tool to underpin water management practices in the region, and hence, to improve water access. This paper presents a machine learning method to map groundwater potential and illustrates it through an application to two regions of Mali. A set of explanatory variables for the presence of groundwater is developed first. Several scaling methods (standardization, normalization, maximum absolute value and min-max scaling) are used to avoid the pitfalls associated with the reclassification of explanatory variables. A number of supervised learning classifiers is then trained and tested on a large borehole database (n = 3,345) in order to find meaningful correlations between the presence or absence of groundwater and the explanatory variables. This process identifies noisy, collinear and counterproductive variables and excludes them from the input dataset. Tree-based algorithms, including the AdaBoost, Gradient Boosting, Random Forest, Decision Tree and Extra Trees classifiers were found to outperform other algorithms on a consistent basis (accuracy > 0.85), whereas maximum absolute value and standardization proved the most efficient methods to scale explanatory variables. Borehole flow rate data is used to calibrate the results beyond standard machine learning metrics, thus adding robustness to the predictions. The southern part of the study area was identified as the better groundwater prospect, which is consistent with the geological and climatic setting. From a methodological standpoint, the outcomes lead to three major conclusions: (1) because there is no aprioristic way to know which algorithm will work better on a given dataset, we advocate the use of a large number of machine learning classifiers, out of which the best are subsequently picked for ensembling; (2) standard machine learning metrics may be of limited value when appraising map outcomes, and should be complemented with hydrogeological indicators whenever possible; and (3) the scaling of the variables helps to minimize bias arising from expert judgement and maintains robust predictive capabilities.

show abstract

“…With the help of ML technology, Islam et al [42] found that higher resolutions of image and elevation detail could have generated more accurate data for environmental risk assessment. Martinez-Santos et al [43] used Random Forest (RF) and Super Vector Machines (SVM) to predict and classify the presence of water bodies. Whilst the algorithm performed well over flat areas, the authors reported some imprecisions over ridges due to the absence of accurate DEMs [43].…”

Section: Introductionmentioning

confidence: 99%

“…Martinez-Santos et al [43] used Random Forest (RF) and Super Vector Machines (SVM) to predict and classify the presence of water bodies. Whilst the algorithm performed well over flat areas, the authors reported some imprecisions over ridges due to the absence of accurate DEMs [43].…”

Section: Introductionmentioning

confidence: 99%

Using Unmanned Aerial Vehicle and LiDAR-Derived DEMs to Estimate Channels of Small Tributary Streams

et al. 2021

View full text Add to dashboard Cite

Defining stream channels in a watershed is important for assessing freshwater habitat availability, complexity, and quality. However, mapping channels of small tributary streams becomes challenging due to frequent channel change and dense vegetation coverage. In this study, we used an Unmanned Aerial Vehicle (UAV) and photogrammetry method to obtain a 3D Digital Surface Model (DSM) to estimate the total in-stream channel and channel width within grazed riparian pastures. We used two methods to predict the stream channel boundary: the Slope Gradient (SG) and Vertical Slope Position (VSP). As a comparison, the same methods were also applied using low-resolution DEM, obtained with traditional photogrammetry (coarse resolution) and two more LiDAR-derived DEMs with different resolution. When using the SG method, the higher-resolution, UAV-derived DEM provided the best agreement with the field-validated area followed by the high-resolution LiDAR DEM, with Mean Squared Errors (MSE) of 1.81 m and 1.91 m, respectively. The LiDAR DEM collected at low resolution was able to predict the stream channel with a MSE of 3.33 m. Finally, the coarse DEM did not perform accurately and the MSE obtained was 26.76 m. On the other hand, when the VSP method was used we found that low-resolution LiDAR DEM performed the best followed by high-resolution LiDAR, with MSE values of 9.70 and 11.45 m, respectively. The MSE for the UAV-derived DEM was 15.12 m and for the coarse DEM was 20.78 m. We found that the UAV-derived DEM could be used to identify steep bank which could be used for mapping the hydrogeomorphology of lower order streams. Therefore, UAVs could be applied to efficiently map small stream channels in order to monitor the dynamic changes occurring in these ecosystems at a local scale. However, the VSP method should be used to map stream channels in small watersheds when high resolution DEM data is not available.

show abstract

Predictive mapping of aquatic ecosystems by means of support vector machines and random forests

Cited by 17 publications

References 39 publications

Shedding light on the decline of Iberian freshwater fish species over the period 1980–2020

Shedding light on the decline of Iberian freshwater fish species over the period 1980–2020

Preprocessing approaches in machine learning-based groundwater potential mapping: an application to the Koulikoro and Bamako regions, Mali

Using Unmanned Aerial Vehicle and LiDAR-Derived DEMs to Estimate Channels of Small Tributary Streams

Contact Info

Product

Resources

About