Identification of Potential Seagrass Areas in the Mediterranean Basin Using Geospatial Artificial Intelligence (GeoAI) Techniques

Şevki Danacıoğlu; Selim Mirsat Gül; Berkay Ballı

Makale detayları

Identification of Potential Seagrass Areas in the Mediterranean Basin Using Geospatial Artificial Intelligence (GeoAI) Techniques

Yazarlar:

Şevki Danacıoğlu | Selim Mirsat Gül | Berkay Ballı

Yayıncı:

Balıkesir University

Ana Başlıklar:

Türkiye Fiziki Coğrafyası | Çevresel Coğrafya

Dergi:

Sosyal Aktörler ve Deneyimler Dergisi Issue 4, pp.5-16

Belge Tipi:

Makale

Yayın Tarihi:

6/2025

Belge Dili:

Türkçe

Document ID (Org No.):

2001345354

pISSN:

3023-6134

Anahtar kelimeler:

Özet:

 In todays world, biodiversity faces many challenges due to global climate change and increasing environmental problems. Seagrass meadows, also referred to as the lungs of the sea, play a crucial role in ecosystems. Seagrasses found in shallow waters reached by sunlight are considered the lungs of the sea due to their capacity to produce oxygen. Additionally, seagrass meadows provide habitats for marine organisms and serve as a food source for grazers. This research aims to predict the habitat areas of seagrass meadows in the Mediterranean using geospatial artificial intelligence GeoAI techniques. GeoAI techniques were employed to determine potential seagrass areas. An artificial intelligence model was developed to create a habitat distribution model considering the conditions of regions where the species exists within the study area. Seven spatial variables influencing the habitat conditions of seagrasses were identified in this study. Data sets from the Copernicus Marine Monitoring Service for the year 2023 were utilized. Data on existing seagrass meadows from the same source were used to train the artificial intelligence model. Within the ArcGIS Pro 3.2 software, the Presence-only Prediction MaxEnt tool was employed to predict the probability of species presence according to the Maximum Entropy method. The cutoff value in the model was set to 0.5, resulting in a 67.7 accuracy in model prediction based on classification results. The resulting map shows areas with significantly high probability values, indicating widespread seagrass habitat presence. These areas include the southern coast of Turkey, the Gulf of Iskenderun, the eastern and western coasts of the Aegean Sea, the southern parts of the Eastern Mediterranean Sea, the coasts of Egypt in the Levantine Sea, the coasts of the Central and Northern Adriatic Sea, the southern part of Sicily, especially the coasts of Tunisia, and the eastern-southeastern coasts of Spain in the Western Mediterranean. The results demonstrate that geospatial artificial intelligence techniques can effectively be used to predict ecosystems and identify terrestrial or marine habitats. The findings of this study are believed to have the potential to guide efforts in determining the distribution and conservation of habitats.

 In todays world, biodiversity faces many challenges due to global climate change and increasing environmental problems. Seagrass meadows, also referred to as the lungs of the sea, play a crucial role in ecosystems. Seagrasses found in shallow waters reached by sunlight are considered the lungs of the sea due to their capacity to produce oxygen. Additionally, seagrass meadows provide habitats for marine organisms and serve as a food source for grazers. This research aims to predict the habitat areas of seagrass meadows in the Mediterranean using geospatial artificial intelligence GeoAI techniques. GeoAI techniques were employed to determine potential seagrass areas. An artificial intelligence model was developed to create a habitat distribution model considering the conditions of regions where the species exists within the study area. Seven spatial variables influencing the habitat conditions of seagrasses were identified in this study. Data sets from the Copernicus Marine Monitoring Service for the year 2023 were utilized. Data on existing seagrass meadows from the same source were used to train the artificial intelligence model. Within the ArcGIS Pro 3.2 software, the Presence-only Prediction MaxEnt tool was employed to predict the probability of species presence according to the Maximum Entropy method. The cutoff value in the model was set to 0.5, resulting in a 67.7 accuracy in model prediction based on classification results. The resulting map shows areas with significantly high probability values, indicating widespread seagrass habitat presence. These areas include the southern coast of Turkey, the Gulf of Iskenderun, the eastern and western coasts of the Aegean Sea, the southern parts of the Eastern Mediterranean Sea, the coasts of Egypt in the Levantine Sea, the coasts of the Central and Northern Adriatic Sea, the southern part of Sicily, especially the coasts of Tunisia, and the eastern-southeastern coasts of Spain in the Western Mediterranean. The results demonstrate that geospatial artificial intelligence techniques can effectively be used to predict ecosystems and identify terrestrial or marine habitats. The findings of this study are believed to have the potential to guide efforts in determining the distribution and conservation of habitats.

 In todays world, biodiversity faces many challenges due to global climate change and increasing environmental problems. Seagrass meadows, also referred to as the lungs of the sea, play a crucial role in ecosystems. Seagrasses found in shallow waters reached by sunlight are considered the lungs of the sea due to their capacity to produce oxygen. Additionally, seagrass meadows provide habitats for marine organisms and serve as a food source for grazers. This research aims to predict the habitat areas of seagrass meadows in the Mediterranean using geospatial artificial intelligence GeoAI techniques. GeoAI techniques were employed to determine potential seagrass areas. An artificial intelligence model was developed to create a habitat distribution model considering the conditions of regions where the species exists within the study area. Seven spatial variables influencing the habitat conditions of seagrasses were identified in this study. Data sets from the Copernicus Marine Monitoring Service for the year 2023 were utilized. Data on existing seagrass meadows from the same source were used to train the artificial intelligence model. Within the ArcGIS Pro 3.2 software, the Presence-only Prediction MaxEnt tool was employed to predict the probability of species presence according to the Maximum Entropy method. The cutoff value in the model was set to 0.5, resulting in a 67.7 accuracy in model prediction based on classification results. The resulting map shows areas with significantly high probability values, indicating widespread seagrass habitat presence. These areas include the southern coast of Turkey, the Gulf of Iskenderun, the eastern and western coasts of the Aegean Sea, the southern parts of the Eastern Mediterranean Sea, the coasts of Egypt in the Levantine Sea, the coasts of the Central and Northern Adriatic Sea, the southern part of Sicily, especially the coasts of Tunisia, and the eastern-southeastern coasts of Spain in the Western Mediterranean. The results demonstrate that geospatial artificial intelligence techniques can effectively be used to predict ecosystems and identify terrestrial or marine habitats. The findings of this study are believed to have the potential to guide efforts in determining the distribution and conservation of habitats.

 Günümüzde canlılık küresel iklim değişikliği ve artan çevre sorunlarına bağlı, birçok sorunla karşı karşıyadır. Denizlerin akciğerleri olarak da isimlendirilen deniz çayırları ekosistem için çok önemli yere sahiptir. Güneş ışınlarının ulaştığı sığ sularda bulunan deniz çayırları, oksijen üretim kapasitesine sahip oldukları için denizlerin akciğerleri konumundadır. Deniz çayırları ayrıca deniz canlıları için yaşam ortamı sağlamakta ve otla beslenen canlıların besin kaynağını oluşturmaktadır. Bu araştırma Akdenizdeki deniz çayırlarının habitat alanlarının coğrafi yapay zeka teknikleriyle tahmin edilmesini amaçlamaktadır. Potansiyel deniz çayırları alanlarının belirlenmesinde Coğrafi Yapay Zeka GeoAI teknikleri kullanılmıştır. Buna göre geliştirilen yapay zeka modeli ile araştırma alanında türün var olduğu bölgelerin koşullarını dikkate alarak bir habitat dağılım modeli oluşturulması hedeflenmektedir. Bu bağlamda çalışmada deniz çayırlarının habitat koşullarını etkileyen 7 mekânsal değişken belirlenmiştir. Araştırmada Copernicus Deniz İzleme Servisi tarafından sağlanan 2023 yılına ait veri setleri kullanılmıştır. Ayrıca aynı kaynak tarafından sağlanan mevcut deniz çayırlarına ait veriler yapay zeka modelinin eğitilmesi amacıyla kullanılmıştır. Araştırmada ArcGIS Pro 3.2 yazılımı içerisinde yer alan ve Maksimum Entropi yöntemine göre türlerin var olma olasılığını tahmin etmek için kullanılan Presence-only Prediction MaxEnt aracı kullanılmıştır. Buna göre modelde kesme değeri Cutoff 0,5 olarak belirlenmiş ve bu bağlamda sınıflandırılma sonuçlarına göre model tahmininde 67.7lik bir doğruluk elde edilmiştir. Elde edilen sonuç haritasında belirgin bir şekilde yüksek olasılık değerlerine sahip olan bölgeler Türkiye güneyinde İskenderun Körfezi, Ege Denizinin Doğu ve Batı kıyıları ile Güney kesimleri, Güney Levanten Denizinde Mısır kıyıları, Orta ve Kuzey Adriyatik Denizi kıyıları, Sicilya Güneyi ile birlikte özellikle Tunus Kıyıları ve Batı Akdenizde İspanyanın Doğu-Güneydoğu kıyıları deniz çayırları habitatlarının yaygın olarak bulunduğu alanları işaret etmektedir. Sonuçlar, coğrafi yapay zeka tekniklerinin, ekosistem tahmininde, karasal veya denizel habitatların belirlenmesinde etkili şekilde kullanılabileceğini göstermektedir. Çalışmanın bulguları, habitatların dağılımını belirleme ve koruma çabalarını yönlendirme potansiyeline sahip olduğu düşünülmektedir.

Makale detayları

Identification of Potential Seagrass Areas in the Mediterranean Basin Using Geospatial Artificial Intelligence (GeoAI) Techniques

Yazarlar:

Şevki Danacıoğlu | Selim Mirsat Gül | Berkay Ballı

Yayıncı:

Balıkesir University

Ana Başlıklar:

Türkiye Fiziki Coğrafyası | Çevresel Coğrafya

Dergi:

Sosyal Aktörler ve Deneyimler Dergisi Issue 4, pp.5-16

Belge Tipi:

Makale

Yayın Tarihi:

6/2025

Belge Dili:

Türkçe

Document ID (Org No.):

2001345354

pISSN:

3023-6134

Geospatial Artificial Intelligence

GeoAI

Habitat Monitoring

Maximum Entropy

Mediterranean BasinAkdeniz Havzası

Coğrafi Yapay Zekâ

GeoAI

Habitat İzleme

Maksimum Entropi

Sayı

Makale detayları

Yazarlar: Şevki Danacıoğlu | Selim Mirsat Gül | Berkay Ballı

Yayıncı: Balıkesir University

Ana Başlıklar: Türkiye Fiziki Coğrafyası | Çevresel Coğrafya

Dergi: Sosyal Aktörler ve Deneyimler Dergisi Issue 4, pp.5-16

Belge Tipi: Makale

Yayın Tarihi: 6/2025

Belge Dili: Türkçe

Document ID (Org No.): 2001345354

pISSN: 3023-6134

Sayı

Yazarlar:

Şevki Danacıoğlu | Selim Mirsat Gül | Berkay Ballı

Yayıncı:

Balıkesir University

Ana Başlıklar:

Türkiye Fiziki Coğrafyası | Çevresel Coğrafya

Dergi:

Sosyal Aktörler ve Deneyimler Dergisi Issue 4, pp.5-16

Belge Tipi:

Makale

Yayın Tarihi:

6/2025

Belge Dili:

Türkçe

Document ID (Org No.):

2001345354

pISSN:

3023-6134