About

Dr. Mutlu Ozdogan is an Associate Professor of Forest and Wildlife Ecology and Environmental Studies at the University of Wisconsin-Madison, affiliated with the Nelson Institute for Environmental Studies. His research focuses on land-use and land-cover conversion, climate change impacts on the global water and energy cycles, and how these impacts interact with ecosystem goods and services that are important to human well-being. He is also interested in improving the information content of satellite observations through algorithm and model development. Dr. Ozdogan has a background in Geological Engineering from Istanbul University, a M.Sc. in Geology from North Carolina State University, and a M.A. in Environmental Remote Sensing from Boston University. He completed his Ph.D. in Geography and Environment at Boston University in 2004, where he worked on water resource scarcity and satellite-assisted methods to search for additional resources in the Middle East. His recent work includes developing a dataset on global irrigation extent using satellite observations, which is used to assess irrigation feedback on climate and the sustainability of agricultural water resources, impacting global food security and human vulnerability. He has been supported by research grants from NASA, NSF, and governments of Oman and the United Arab Emirates.

Research topics

• Computer Science
• Machine Learning
• Artificial Intelligence
• Agronomy
• Geography
• Environmental science
• Statistics
• Mathematics
• Remote sensing
• Environmental resource management
• Algorithm
• Database

Selected publications

• Data for: Impacts of logging, hunting, and conservation on vocalizing biodiversity in Gabon

  DRYAD · 2026-05-13

  datasetOpen access

  Tropical forests support two-thirds of the world's biodiversity, contribute to global climate regulation, and support the culture and livelihoods of forest-dependent people. Much of extant tropical forest is subject to selective logging and hunting - extractive activities that potentially alter ecosystem function and species diversity. However, the collective impact of these threats, especially in the context of protected vs unprotected areas, is not fully understood. Here we assess how vocalizing biodiversity responds to logging and hunting, across the diel cycle, seasonally, and between protected and unprotected landscapes in Gabon. We compared soundscape saturation across 109 sites in national parks, Forest Stewardship Council (FSC) certified, and non-certified logging concessions. We estimated hunting pressure by quantifying gunshots and relative accessibility per site. Overall, we found that the soundscapes of FSC-certified concessions resembled national parks (selectively logged 20+ years ago) more so than non-certified concessions. We also found that never logged sites, part of a proposed community conserved area, had different soundscapes than all other categories, including national parks. Unlogged sites had higher saturation than logging concessions at dusk and dawn. Soundscapes and hunting pressure were highly variable across different concessions. We found that higher gunshot rates and recent logging were associated with lower soundscape saturation overall. Based on our findings, we recommend that (i) the very few never logged forests that remain (and are not yet protected) should be urgently withdrawn from selective logging, and (ii) FSC or other certification schemes should be promoted in Gabon, with an emphasis on sustainable hunting.

  PublisherDOI

• The Peak Value Method (PVM): A Phenology-Guided, Training-Independent Approach for Large-Area Time-Series Winter Wheat Mapping in the Cross-Border Region Between Bulgaria and Türkiye

  IEEE Transactions on Geoscience and Remote Sensing · 2026-01-01

  articleOpen access

  Accurate, field-scale mapping of crop-type areas across large and diverse regions remains challenging due to the need for extensive training data, inconsistent ground surveys, and the computational demands of supervised methods. This study presents the Peak Value Method (PVM)—a dynamic, phenology-guided, and training-independent classification framework for mapping winter wheat extent over 20 years (2001–2020) across the socioeconomically distinct but agroecologically similar cross-border region between Bulgaria and Türkiye. By leveraging Landsat-derived Enhanced Vegetation Index (EVI) time series, PVM uses region-specific, empirically optimized thresholds for peak vegetation intensity (MaxEVI) and its timing (MaxDOY) aligned with local cropping calendars, to classify wheat pixels at 30-meter resolution. The method achieved strong validation performance, with overall accuracies ranging from 86% to 98.5% and an average <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">R</i><sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> of 0.85, based on comparisons with official wheat statistics and ground-collected data. Final thresholds were calibrated through a sensitivity analysis of MaxEVI and MaxDOY, ensuring regional adaptability. PVM incorporated annual cropland masks, generated using a generalized classifier (∼88% accuracy), which reduced wheat commission errors by ∼30%. Analysis of spatiotemporal dynamics showed distinct trends in the cross-border regions. This study delivers the first consistent, high-resolution, two-decade winter wheat maps for this region and provides a scalable and reproducible solution for long-term operational agricultural monitoring without training data dependency. All data and code are openly available at https://doi.org/10.5281/zenodo.15612128.

  PublisherDOI

• The Peak Value Method (PVM): A Phenology-Guided, Training-Independent Approach for Large-Area Time-Series Winter Wheat Mapping

  2025-07-22 · 1 citations

  preprintOpen access

  Accurate, field-scale mapping of crop-type areas across large and diverse regions remains challenging due to the need for extensive training data, inconsistent ground surveys, and the computational demands of supervised methods. This study presents the Peak Value Method (PVM)-a dynamic, phenologyguided, and training-independent classification framework for mapping winter wheat extent over 20 years (2001-2020) across the socioeconomically distinct but agroecologically similar crossborder region between Bulgaria and T ürkiye. By leveraging Landsat-derived Enhanced Vegetation Index (EVI) time series, PVM uses region-specific, empirically optimized thresholds for peak vegetation intensity (MaxEVI) and its timing (MaxDOY) aligned with local cropping calendars, to classify wheat pixels at 30-meter resolution. The method achieved strong validation performance, with overall accuracies ranging from 86% to 98.5% and an average R 2 of 0.85, based on comparisons with official wheat statistics and ground-collected data. Final thresholds were calibrated through a sensitivity analysis of MaxEVI and MaxDOY, ensuring regional adaptability. PVM incorporated annual cropland masks, generated using a generalized classifier (∼88% accuracy), which reduced wheat commission errors by ∼30%. Analysis of spatiotemporal dynamics showed distinct trends in the cross-border regions. This study delivers the first consistent, high-resolution, two-decade winter wheat maps for this region and provides a scalable and reproducible solution for long-term operational agricultural monitoring without training data dependency. All data and code are openly available at https://doi.org/10.5281/zenodo.15612128.

  PublisherOA PDFDOI

• Field-Scale Rice Area and Yield Mapping in Sri Lanka with Optical Remote Sensing and Limited Training Data

  Remote Sensing · 2025-09-03 · 1 citations

  articleOpen access1st authorCorresponding

  Rice is a staple crop for over half the world’s population, and accurate, timely information on its planted area and production is crucial for food security and agricultural policy, particularly in developing nations like Sri Lanka. However, reliable rice monitoring in regions like Sri Lanka faces significant challenges due to frequent cloud cover and the fragmented nature of smallholder farms. This research introduces a novel, cost-effective method for mapping rice-planted area and yield at field scales in Sri Lanka using optical satellite data. The rice-planted fields were identified and mapped using a phenologically tuned image classification algorithm that highlights rice presence by observing water occurrence during transplanting and vegetation activity during subsequent crop growth. To estimate yields, a random forest regression model was trained at the district level by incorporating a satellite-derived chlorophyll index and environmental variables and subsequently applied at the field level. The approach has enabled the creation of two decades (2000–2022) of reliable, field-scale rice area and yield estimates, achieving map accuracies between 70% and over 90% and yield estimates with less than 20% error. These highly granular results, which are not available through traditional surveys, show a strong correlation with government statistics. They also demonstrate the advantages of a rule-based, phenology-driven classification over purely statistical machine learning models for long-term consistency in dynamic agricultural environments. This work highlights the significant potential of remote sensing to provide accurate and detailed insights into rice cultivation, supporting policy decisions and enhancing food security in Sri Lanka and other cloud-prone regions.

  PublisherOA PDFDOI

• Field-Scale Rice Area and Yield Mapping in Sri Lanka with Optical Remote Sensing and Limited Training Data

  Washington, DC: World Bank eBooks · 2025-08-21

  bookOpen access1st authorCorresponding

  Rice is a staple crop for over half the world’s population, and accurate, timely information on its planted area and production is crucial for food security and agricultural policy, particularly in developing nations like Sri Lanka. However, reliable rice monitoring in regions like Sri Lanka faces significant challenges due to frequent cloud cover and the fragmented nature of small-holder farms. This research introduces a novel, cost-effective method for mapping rice planted area and yield at field scales in Sri Lanka using optical satellite data. The rice planted fields were identified and mapped using a phenologically-tuned image classification algorithm that high-lights rice presence by observing water occurrence during transplanting and vegetation activity during subsequent crop growth. To estimate yields, a random forest regression model was trained at the district level by incorporating a satellite-derived chlorophyll index and environmental variables and subsequently applied at the field level. The approach has enabled the creation of two decades (2000–2022) of reliable, field-scale rice area and yield estimates, achieving map accuracies between 70% and over 90% and yield estimations with less than 20% RMSE. These highly granular results, which were previously unattainable through traditional surveys, show strong correlation with government statistics. They also demonstrate the ad-vantages of a rule-based, phenology-driven classification over purely statistical machine learning models for long-term consistency in dynamic agricultural environments. This work highlights the significant potential of remote sensing to provide accurate and detailed insights into rice cultivation, supporting policy decisions and enhancing food security in Sri Lanka and other cloud-prone regions.

  PublisherOA PDFDOI

• Global Food Security Support Analysis Data (GFSAD) Using Remote Sensing in Support of Food and Water Security in the 21st Century

  2024-10-30

  book-chapter

  The overarching goal of this chapter is to provide a comprehensive overview of the state-of-art of global cropland mapping procedures using remote sensing as characterized and envisioned by the “Global Food Security Support Analysis Data @ 30 m (GFSAD30)” project working group team. First, the chapter will provide an overview of existing cropland maps and their characteristics along with establishing the gaps in knowledge related to global cropland mapping. Second, definitions of cropland mapping along with key parameters involved in cropland mapping based on their importance in food security analysis, and cropland naming conventions for standardized cropland mapping using remote sensing will be presented. Third, existing methods and approaches for cropland mapping will be discussed. This will include the type of remote sensing data used in cropland mapping and their characteristics along with discussions on the secondary data, field-plot data, and cropland mapping algorithms. Fourth, currently existing global cropland products derived using remote sensing will be presented and discussed. Fifth, a synthesis of all existing products leading to a composite global cropland extent version 1.0 (GCE V1.0) is presented and discussed. Sixth, a way forward for advanced global cropland mapping is visualized.

  PublisherDOI

• A Machine Learning and Remote Sensing‐Based Model for Algae Pigment and Dissolved Oxygen Retrieval on a Small Inland Lake

  Water Resources Research · 2024-02-28 · 40 citations

  articleOpen access

  Abstract Excessive algae growth can lead to negative consequences for ecosystem function, economic opportunity, and human and animal health. Due to the cost‐effectiveness and temporal availability of satellite imagery, remote sensing has become a powerful tool for water quality monitoring. The use of remotely sensed products to monitor water quality related to algae and cyanobacteria productivity during a bloom event may help inform management strategies for inland waters. To evaluate the ability of satellite imagery to monitor algae pigments and dissolved oxygen conditions in a small inland lake, chlorophyll‐a, phycocyanin, and dissolved oxygen concentrations are measured using a YSI EXO2 sonde during Sentinel‐2 and Sentinel‐3 overpasses from 2019 to 2022 on Lake Mendota, WI. Machine learning methods are implemented with existing algorithms to model chlorophyll‐a, phycocyanin, and Pc:Chla. A novel machine learning‐based dissolved oxygen modeling approach is developed using algae pigment concentrations as predictors. Best model results based on Sentinel‐2 (Sentinel‐3) imagery achieved R 2 scores of 0.47 (0.42) for chlorophyll‐a, 0.69 (0.22) for phycocyanin, and 0.70 (0.41) for Pc:Chla. Dissolved oxygen models achieved an R 2 of 0.68 (0.36) when applied to Sentinel‐2 (Sentinel‐3) imagery, and Pc:Chla is found to be the most important predictive feature. Random forest models are better suited to water quality estimations in this system given built in methods for feature selection and a relatively small data set. Use of these approaches for estimation of Pc:Chla and dissolved oxygen can increase the water quality information extracted from satellite imagery and improve characterization of algae conditions among inland waters.

  PublisherOA PDFDOI

• A new remote sensing method to map field- scale time-series wheat without training data in Bulgaria

  2024-07-07

  article

  Accurate and up-to date wheat area estimate is a critical input for food security and policy related studies. Wheat as a staple crop covers fifty percent of agricultural lands in Bulgaria. In Bulgaria, current wheat area statistics is available at regional scale (equivalent to census regions in the United States) until year 2017 from government departments. Remote sensing provides extraordinary capability to map field-scale temporally consistent wheat area but, traditional large area mapping methods using remote sensing require large volume of training data and large computing resources. This study presents a new approach called "Peak Value Method" (PVM) to automatically map the wheat cultivated areas in Bulgaria without training data. This study mapped wheat areas between 2001 and 2020 using PVM through Landsat derived enhanced vegetation index (EVI) time series data and phenological understanding of crop. To our knowledge, this is the first study to map wall to wall high resolution time-series wheat areas with reliable accuracy assessment in this study area. This study offers three main outputs: 1. a new, simple but effective remote sensing method to map wheat crop without training data in the study area, 2. first field scale (30-m) time series (2001-2020) wheat maps with reliable accuracy covering Northcentral, Northeast, Southcentral and Southeast regions of Bulgaria and, 3. fills the gap in the knowledge about field-scale wheat information in the study area and it’s accessibility through remote sensing to study the environment related questions.

  PublisherDOI

• Image Classification Methods in Land Cover and Land Use

  2024-10-29

  book-chapter1st authorCorresponding

  Today, information on land cover and land use is almost exclusively derived from remotely sensed observations at various spatial and temporal scales. The advantages of these observations include, but not limited to, synoptic view, availability of spectral bands that help distinguish land surface properties, archived temporal record, and digital nature. The principal form of deriving land cover information from remotely sensed images is classification. In the context of remote sensing, classification refers to the process of translating observations into land cover categories with clearly defined biogeophysical function. For example, a typical land cover map may contain categories like forest, water, agriculture, and so on. These maps are then used in a growing number of environmental applications, from resource management to global change studies. To this end, the purpose of this chapter is to review existing and emerging image classification methods applied to remote sensing.

  PublisherDOI

• Impacts of Logging, Hunting, and Conservation on Vocalizing Biodiversity in Gabon

  SSRN Electronic Journal · 2024-01-01 · 1 citations

  preprintOpen access
  PublisherDOI

Recent grants

• Belmont Forum - European Joint Programming Initiative Collaborative Research: Hydro-social and environmental impacts of sugar cane production on land use and food security

  NSF · $26k · 2014–2017

Frequent coauthors

• Curtis E. Woodcock

  14 shared

• Yanghui Kang

  University of California, Berkeley

  14 shared

• Prasad S. Thenkabail

  Astrogeology Science Center

  13 shared

• Martha C. Anderson

  13 shared

• Benjamin F. Zaitchik

  Planetary Science Institute

  12 shared

• Guido D. Salvucci

  10 shared

• Kamini Yadav

  8 shared

• Aparna Phalke

  University of Alabama in Huntsville

  8 shared

Labs

• Forest and Wildlife EcologyPI

Education

• Ph.D., Environmental Science

  University of Wisconsin–Madison

  2004

• M.S., Environmental Science

  University of Wisconsin–Madison

  2001

• B.S., Environmental Science

  University of Wisconsin–Madison

  1998