Abstract: The experiments were conducted in the artificial climate chamber from September to November 2021, using "Hongyan" strawberry(Fragaria×ananassa Duch "Benihope") as the test material for dynamic high- temperature environmental control experiments at the seedling stage (9−12 true leaves, leaf length ≥ 5 cm), with a daily maximum temperature of 32°C as the starting point and four levels of daily maximum/daily minimum temperatures of 32°C/22°C, 35°C/25°C, 38°C/28°C, and 41°C/31°C for 2d, 5d, 8d, and 11d, respectively, and 28°C/18°C as the control (CK).The relative air humidity is 60%−70%, the photoperiod is 12h/12h (6:00−18:00), and the light intensity is 800μmol·m⁻²·s⁻¹. The chlorophyll content and hyperspectral reflectance of leaves under different treatments were measured, and the original spectrum was changed to refine the spectral characteristic information. Based on the correlation analysis, the original and first-order sensitive band vegetation indices were established, and then the spectral characteristic parameters that characterize chlorophyll content were screened, in order to build the best estimation model of chlorophyll content. The results showed that: (1) the chlorophyll a, chlorophyll b and total chlorophyll (a+b) contents of strawberry leaves tended to decrease as the temperature increased and the duration of high temperature increased. (2) The original spectra of strawberry leaves had green peaks and red valleys in the visible region, except for the green peaks and red valleys, which did not differ significantly among treatments, and the reflectance in the NIR region under high temperature conditions showed different degrees of increase compared with CK. Compared with the original spectra, the first-order derivative spectral curves oscillated more strongly and were able to highlight the red edge parameters significantly. The red edge position λ_r of each treatment was stable at 716nm, and the difference between the red edge amplitude D_r and the red edge area S_r was significant; while the green peak (near 550 nm) and red valley (near 675nm) of each treatment were completely highlighted in the continuous uniform removal spectra. (3) Based on the correlation analysis between spectral reflectance and chlorophyll content, R₇₄₇, R₈₀₀ and R'₇₁₆, R'₉₀₆, which have the strongest correlation between the original and first derivative spectra in the visible and near-infrared bands, were selected as the combination of sensitive bands to construct the vegetation index. (4) PVI, MSAVI, TSAVI, TSAVI', DVI', MSAVI', PVI', SAVI', Dr and Sr indices correlated with chlorophyll content at highly significant levels and can be used as hyperspectral characteristic parameters to characterize the chlorophyll content of greenhouse strawberry leaves in response to high temperature stress at seedling stage. The stepwise regression model with TSAVI, DVI' and PVI' vegetation indices was the best estimation model for chlorophyll content, with a coefficient of determination (R²) of 0.843, root mean square error (RMSE) of 0.379 and relative error (RE) of 12.65%.

Key words: Strawberry, High temperature stress, Hyperspectral parameters, Chlorophyll content, Estimation model