We combine farm accounting data with high-resolution meteorological data, and climate scenarios to estimate climate change impacts and adaptation potentials at the farm level. To do so, we adapt the seminal model of Moore and Lobell (2014) who applied panel data econometrics to data aggregated from the farm to the regional (subnational) level. We discuss and empirically investigate the advantages and challenges of applying such models to farm-level data, including issues of endogeneity of explanatory variables, heterogeneity of farm responses to weather shocks, measurement errors in meteorological variables, and aggregation bias. Empirical investigations into these issues reveal that endogeneity due to measurement errors in temperature and precipitation variables, as well as heterogeneous responses of farms toward climate change may be problematic. Moreover, depending on how data are aggregated, results differ substantially compared to farm-level analysis. Based on data from Austria and two climate scenarios (Effective Measures and High Emission) for 2040, we estimate that the profits of farms will decline, on average, by 4.4% (Effective Measures) and 10% (High Emission). Adaptation options help to considerably ameliorate the adverse situation under both scenarios. Our results reinforce the need for mitigation and adaptation to climate change.

中文翻译：

---

利用农场层面的面板数据来估计气候变化的影响和适应潜力

我们将农场核算数据与高分辨率气象数据和气候情景相结合，以估计农场层面的气候变化影响和适应潜力。为此，我们采用了 Moore 和 Lobell（2014 年）的开创性模型，他们将面板数据计量经济学应用于从农场到区域（国家以下）级别的数据汇总。我们讨论并实证研究了将此类模型应用于农场数据的优势和挑战，包括解释变量的内生性问题、农场对天气冲击反应的异质性、气象变量的测量误差和汇总偏差。对这些问题的实证研究表明，由于温度和降水变量的测量误差以及农场对气候变化的异质反应导致的内生性可能存在问题。此外，根据数据的聚合方式，结果与农场层面的分析有很大差异。根据奥地利的数据和 2040 年的两种气候情景（有效措施和高排放），我们估计农场的利润将平均下降 4.4%（有效措施）和 10%（高排放）。适应方案有助于大大改善这两种情况下的不利情况。我们的结果强化了减缓和适应气候变化的必要性。适应方案有助于大大改善这两种情况下的不利情况。我们的结果强化了减缓和适应气候变化的必要性。适应方案有助于大大改善这两种情况下的不利情况。我们的结果强化了减缓和适应气候变化的必要性。